WO2014076844A1 - Autonomous movement system and control device - Google Patents

Autonomous movement system and control device Download PDF

Info

Publication number
WO2014076844A1
WO2014076844A1 PCT/JP2012/079970 JP2012079970W WO2014076844A1 WO 2014076844 A1 WO2014076844 A1 WO 2014076844A1 JP 2012079970 W JP2012079970 W JP 2012079970W WO 2014076844 A1 WO2014076844 A1 WO 2014076844A1
Authority
WO
WIPO (PCT)
Prior art keywords
map
coordinates
correction
autonomous mobile
travel locus
Prior art date
Application number
PCT/JP2012/079970
Other languages
French (fr)
Japanese (ja)
Inventor
大島　章
山本　健次郎
一野瀬　亮子
幸彦 小野
Original Assignee
株式会社日立製作所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社日立製作所 filed Critical 株式会社日立製作所
Priority to JP2014546831A priority Critical patent/JP5930346B2/en
Priority to PCT/JP2012/079970 priority patent/WO2014076844A1/en
Publication of WO2014076844A1 publication Critical patent/WO2014076844A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0268—Control of position or course in two dimensions specially adapted to land vehicles using internal positioning means
    • G05D1/0272—Control of position or course in two dimensions specially adapted to land vehicles using internal positioning means comprising means for registering the travel distance, e.g. revolutions of wheels
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0268—Control of position or course in two dimensions specially adapted to land vehicles using internal positioning means
    • G05D1/0274—Control of position or course in two dimensions specially adapted to land vehicles using internal positioning means using mapping information stored in a memory device
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0276—Control of position or course in two dimensions specially adapted to land vehicles using signals provided by a source external to the vehicle
    • G05D1/0278—Control of position or course in two dimensions specially adapted to land vehicles using signals provided by a source external to the vehicle using satellite positioning signals, e.g. GPS

Definitions

  • the present invention relates to a technology of an autonomous movement system and a control device that estimates a self-position based on environmental information and a map and autonomously moves based on the estimated self-position.
  • Realization of an autonomous mobile device requires a system equipped with a self-position estimation technique for accurately grasping where the autonomous mobile device itself is located and not making a mistake in the traveling position.
  • environmental information such as image information, shape information, and coordinates prepared in advance is used as a map for self-position estimation (hereinafter also simply referred to as a map).
  • the autonomous mobile device identifies the position of the autonomous mobile device itself on the map by comparing environmental information acquired from the surroundings of the autonomous mobile device itself with a map prepared in advance. For this reason, self-position estimation accuracy improves by generating a map with high accuracy, and the autonomous mobile device can perform stable autonomous movement.
  • it is desirable that the information held in the map is acquired from the autonomous mobile device.
  • maps generated based on environmental information acquired from the sky using satellites and aircraft have a much different environment appearance than autonomous mobile devices that move on the ground (floor surface or road surface). It cannot be used as a map for moving.
  • a map generated based on environmental information acquired from a distant point even at the same height is not likely to be used as a map because it is likely to look different from the autonomous mobile device. For this reason, it is common to manually run an autonomous mobile device (maneuvered by a person) in advance in an environment subject to autonomous movement, acquire surrounding environment information, and generate a map based on this environment information. Has been done. By doing in this way, it becomes possible to generate a map with high certainty in collation during autonomous movement.
  • a map generation method for accurately connecting and integrating maps obtained by dividing into a plurality of maps is disclosed (for example, see Patent Document 1 and Patent Document 2).
  • the technique described in Patent Document 1 extracts one or more feature points from the environment, derives a constraint relationship by associating the feature points between the divided maps, and generates one wide-area map based on this. is doing.
  • Patent Document 1 acquires and divides an autonomously moving environment as a partial map without duplication, and determines and corrects the mutual positional relationship based on the constraint relationship of each feature information. By doing so, one map is generated.
  • the partial maps are set for each narrow area having a characteristic environmental shape such as an intersection or a corner.
  • the constraint relationship is configured such that the map is generated by the user inputting the positional relationship of the adjacent partial maps, the parallelism of the shape, or the like based on the design drawing of the surrounding environment.
  • the constraint relationship refers to a relationship in which positions are constrained between a plurality of pieces of information in order to link the locations between different environmental information including the same location.
  • different environmental information including the same point includes, for example, environmental information of the same point in different partial maps, environmental information of the same point acquired by different routes and means in one or more partial maps, etc. This is called a restraint relationship.
  • the constraint means a relationship such as position / posture between two or more pieces of information.
  • the coordinates on one map are It shows what kind of positional relationship it is with respect to which coordinate on the other map.
  • a constraint between GPS (Global Positioning System) positioning information and movement coordinates such as odometry is also referred to as a constraint relationship as to which point a certain movement coordinate is in the absolute coordinate system in the GPS positioning information.
  • Patent Document 2 is based on the screen output provided in the autonomous mobile device and the interface operated by a human by button input, thereby associating connection points between a plurality of maps into one wide area map. Is generated.
  • the technology described in Patent Document 2 acquires an environment in which autonomous movement is performed as a partial map in the same manner as the technology described in Patent Document 1.
  • the partial maps overlap each other little by little, and one wide-area map is generated by the user setting the constraint relationship between the positions and orientations of adjacent partial maps by screen output or button input.
  • the autonomous mobile device is configured to change partial maps and move to a target point.
  • the technique described in Patent Document 2 when there is a change in the shape of the environment or the travel route, only the partial map including the point is acquired again and the map is generated again.
  • the technique described in Patent Document 1 makes it difficult to extract feature information such as parallelism because there are fewer straight portions when there is a complex-shaped structure, making it difficult to use the feature information.
  • the technique described in Patent Document 1 causes a situation where the partial maps cannot be accurately connected. Therefore, the technique described in Patent Document 1 cannot be applied to wide-area map creation when there is a complex-shaped structure.
  • the technique described in Patent Document 1 is based on the premise that the road surface of the environment is horizontal, and does not take into account the case of having a slope or a three-dimensional intersection. Therefore, in the technique described in Patent Document 1, if the environment information includes three-dimensional information, a shift (error) occurs in the height direction in the generated map.
  • the technique described in Patent Document 2 distortion remains in each partial map.
  • the distortion is an error caused by a measurement error such as odometry.
  • the technique described in Patent Document 2 can be used for a wide-area map generation because a simple constraint relationship may be defined as long as it is a line-like traveling locus.
  • the technique described in Patent Document 2 needs to further subdivide the partial map and set a large number of constraint relationships. is there.
  • the technique described in Patent Document 2 cannot be applied to wide-area map generation for dealing with a mesh-like traveling locus.
  • the technique described in Patent Document 2 is based on the premise that the road surface of the environment is horizontal as in the technique described in Patent Document 1, and the case where the environment information such as slopes and three-dimensional intersections has three-dimensional information. Not considered.
  • the present invention has been made in view of such a background, and an object of the present invention is to improve the certainty in connection of partial maps.
  • the present invention performs a first correction for deforming a travel locus based on absolute coordinates, and further detects an overlapping travel region in which the corrected travel tracks are close to each other. After associating the relative coordinates on the travel locus in the travel region, a second correction for further correcting the travel locus is performed.
  • FIG. 10 is a diagram (part 1) illustrating a conceptual diagram of determination of availability of environmental information acquired from the sky. It is FIG. (2) which shows the conceptual diagram of the availability determination of the environmental information acquired from the sky. It is a flowchart which shows the procedure of the connection of the partial map which concerns on this embodiment. It is a figure for demonstrating the connection method of the partial map which concerns on this embodiment.
  • the autonomous mobile system according to the present embodiment makes it possible to accurately derive the movement coordinates by correcting the movement coordinates (movement locus) when the autonomous movement apparatus is manually driven, including the height direction. .
  • the autonomous mobile system according to this embodiment generates a map for self-position estimation by accurately connecting and integrating shape information (partial maps) obtained by dividing even if the environment of autonomous movement is wide area By enabling the technology, a highly accurate self-position estimation technology is realized.
  • the partial map is a map related to a partial region obtained by dividing the region where the autonomous mobile device travels, and is set so that overlapping regions (inter-map overlapping regions) are generated between the partial maps.
  • the inter-map overlap area will be described later with reference to FIG.
  • FIG. 1 is a diagram illustrating a configuration example of an autonomous mobile system according to the present embodiment.
  • the autonomous mobile system 1 includes an environment information acquisition unit (environment information acquisition unit) 101, a travel locus correction unit (travel locus correction unit) 102, a partial map generation unit (map generation unit) 103, and a map generation unit (map generation unit) 104.
  • the environment information acquisition unit 101 is configured by combining a plurality of measurement units.
  • the environment information acquisition unit 101 includes, for example, a laser distance sensor, a monocular or compound eye camera system, a GPS sensor, an atmospheric pressure sensor, and the like.
  • laser survey information may be acquired from an image acquired from an aircraft or a satellite, or a laser surveying device (not shown) provided outside the autonomous mobile device 10.
  • the environment information acquisition unit 101 may acquire information manually surveyed separately.
  • the environment information acquisition unit 101 may be combined with a device that acquires information inside the autonomous mobile device 10 in addition to a device that acquires external information from the autonomous mobile device 10.
  • an environment information acquisition unit 101 for example, it is possible to acquire movement amount information that can be acquired from a moving mechanism such as a wheel, crawler, or leg for movement of the autonomous mobile device 10, and momentum information such as angular velocity or acceleration.
  • a moving mechanism such as a wheel, crawler, or leg for movement of the autonomous mobile device 10
  • momentum information such as angular velocity or acceleration.
  • movement coordinates There are inertial measurement sensors. Movement amount information that can be acquired from movement mechanisms such as wheels, crawlers, and legs, and momentum information such as angular velocity and acceleration are referred to as movement coordinates.
  • Each environment information acquired by the environment information acquisition unit 101 (GPS coordinates by the GPS sensor, surrounding shape information by the laser distance sensor, shape information acquired from the aircraft, etc.) is associated with the movement coordinates and stored in the environment information. Stored in the unit 121.
  • the environment information acquisition unit 101 includes a relative coordinate that indicates the relative position of the autonomous mobile device during movement, an absolute coordinate that indicates the absolute position of the autonomous mobile device during movement, and environmental information of the surrounding area during movement. And get
  • the environment information storage unit 121 also stores environment information (manual environment information) obtained when the autonomous mobile device 10 is steered and traveled by the user, and the autonomous mobile device 10 moves autonomously.
  • the environment information (autonomous environment information) obtained when doing so is also stored.
  • the self-position estimation unit 105 described later is used for estimation of the self-position, or the path generation unit 106 described later is used for arrangement of surrounding structures and road surfaces. It is used when grasping the situation and moving an appropriate route.
  • the travel locus correction unit 102 corrects the travel locus (described later in detail) based on a plurality of types of environment information acquired by the environment information acquisition unit 101.
  • the partial map generation unit 103 generates a partial map by complementing the shape data included in the environment information based on the travel locus corrected by the travel locus correction unit 102.
  • the partial map generation unit 103 may generate a partial map using only manual environment information, or may generate a partial map using both manual environment information and autonomous environment information.
  • the partial map generation unit 103 may generate a partial map by directly using environmental information from a device that acquires environmental information provided outside the autonomous mobile device 10 (not shown).
  • the map generation unit 104 connects a plurality of partial maps generated by the partial map generation unit 103 to generate a map for self-position estimation.
  • the map generation unit 104 may generate / update a map online when the autonomous mobile device 10 moves, or may generate / update a map offline.
  • the generated map is stored in the map storage unit 122.
  • the self-position estimation unit 105 estimates the self-position by matching (matching) the environment information acquired during autonomous movement with the map stored in the map storage unit 122.
  • a self-position estimation technique in addition to the above-described technique, self-position estimation based on accumulation of information using the vehicle body movement amount or other internal information may be used, or self-position estimation based on GPS positioning information is used. May be. Further, by combining these self-position estimation methods and applying a filtering process (for example, Kalman filtering or an application method thereof), the self-position estimation results by the respective methods may be merged. In association with the map, techniques used for image processing and point cloud processing are used.
  • the route generation unit 106 extracts surrounding obstacles, road surface steps, pedestrian positions, and pedestrian movement speeds from environmental information acquired during autonomous movement, and performs graph search processing and motion model simulation processing. By using it, the moving direction and speed of the autonomous mobile device 10 are determined.
  • the movement control unit 107 moves the autonomous mobile device 10 based on the moving direction and speed of the autonomous mobile device 10 determined by the route generation unit 106. Thereby, the autonomous mobile device 10 automatically moves from one point to another target point.
  • the environment information acquisition unit 101, the self-position estimation unit 105, the route generation unit 106, the movement control unit 107, the environment information storage unit 121, and the map storage unit 122 may be mounted on the autonomous mobile device 10. Conceivable. Similarly, it is conceivable that the environment information storage unit 121, the travel locus correction unit 102, the partial map generation unit 103, the map generation unit 104, and the map storage unit 122 are mounted on the control device 20 installed in a control station or the like. .
  • the units 101 to 107, 121, and 122 are not necessarily mounted on the autonomous mobile device 10 and the control device 20 as shown in FIG. 1.
  • all the units 101 to 107, 121, and 122 are all mounted on the autonomous mobile device 10.
  • the environment information acquisition unit 101, the travel locus correction unit 102, the self-position estimation unit 105, the route generation unit 106, the movement control unit 107, and the environment information storage unit 121 are mounted on the autonomous mobile device 10. Also good.
  • the environmental information acquired by the autonomous mobile device 10 is transmitted to the control device 20. Then, after the control device 20 corrects the travel locus based on the environmental information, a partial map is generated, and the partial map is further connected to generate a wide-area map. Then, the control device 20 sends the generated map to the autonomous mobile device 10.
  • the autonomous mobile device 10 estimates its own position based on the sent map and moves.
  • FIG. 2 is a diagram illustrating a hardware configuration example of the autonomous mobile device and the control device according to the present embodiment.
  • FIG. 2A is a diagram illustrating a hardware configuration example of the autonomous mobile device 10.
  • a CPU Central Processing Unit
  • a memory 202 such as a ROM (Read Only Memory)
  • a communication interface 203 is connected to each other via a bus 204.
  • a program is stored in the memory 202, and when the CPU 201 executes the program, the self-position estimation unit 105, the route generation unit 106, the movement control unit 107, and the like of FIG.
  • FIG. 2B is a diagram illustrating a hardware configuration example of the control device 20.
  • a CPU 211 a CPU 211, a RAM (Random Access Memory) 212, a ROM 213, a communication interface 214, and an HD (Hard Disk) 215 are connected to each other via a bus 216.
  • a program stored in the ROM 213 and the HD 215 is expanded in the RAM 212, and the CPU 211 executes the program, thereby realizing the travel locus correction unit 102, the partial map generation unit 103, the map generation unit 104, and the like.
  • FIG. 3 is a flowchart showing the procedure of map generation processing according to the present embodiment. This process is a process performed during manual travel.
  • the environment information acquisition unit 101 acquires environment information related to a partial area of the traveling road (S101).
  • the travel locus correction unit 102 performs a travel locus correction process for correcting the travel locus based on the movement coordinates, which is one of the environment information (S102). Step S102 will be described later with reference to FIGS.
  • the moving coordinates are coordinate data acquired by dead reckoning, that is, wheel odometry, gyro odometry, visual odometry (hereinafter simply referred to as odometry).
  • trajectory is a locus
  • the partial map generation unit 103 generates a partial map by complementing the shape data obtained from the environment information based on the corrected travel locus (S103).
  • the map generation unit 104 connects the partial maps based on the corrected travel locus, generates a wide area map (S104), and stores the map in the map storage unit 122. Step S104 will be described later with reference to FIGS.
  • FIG. 4 is a flowchart showing the procedure of the autonomous movement process according to the present embodiment.
  • the environment information acquisition unit 101 acquires environment information (S201).
  • the self-position estimation unit 105 estimates the self-position by collating the environment information acquired in step S201 with the map stored in the map storage unit 122 (S202).
  • the route traveled by the autonomous mobile device 10 based on the stationary obstacle, the moving obstacle information, the self-location information, the map, and the like extracted from the environment information acquired in step S201 by the route generation unit 106. Is generated (S203).
  • the movement control unit 107 moves the autonomous mobile device 10 according to the generated route (S204).
  • FIG. 5 is a diagram showing a detailed procedure of the travel locus correction process (step S102 in FIG. 3) according to the present embodiment.
  • the travel locus correction unit 102 performs first correction, which is correction of the travel locus based on the movement coordinates, using GPS positioning information that is one of the environment information (S301).
  • the movement coordinates are obtained by integrating movement amount information and momentum information (pointing to angular velocity, acceleration, and geomagnetic direction).
  • the movement coordinates accumulate errors due to slippage between the movement mechanism and the road surface, sensor measurement error, and the like. That is, in the movement coordinates, the error increases as the movement distance increases.
  • the trajectory connecting the movement coordinates is the travel trajectory.
  • FIG. 6 is a diagram illustrating an example of a travel locus based on movement coordinates according to the present embodiment.
  • a travel locus 300 in FIG. 6 in the present embodiment, a travel locus based on three-dimensional movement coordinates having information in the height direction based on three-dimensional odometry is assumed.
  • the travel locus 300 in FIG. 6 does not consider errors due to odometry, which will be described later.
  • the travel locus 300 actually includes errors as described later. Therefore, the traveling locus correction unit 102 corrects the traveling locus in step S301 in FIG. 5 to three-dimensionally correct the horizontal direction and the height direction of the movement coordinates.
  • the moving coordinates are not limited to the three-dimensional coordinates as shown in FIG. 6, and may be two-dimensional coordinates based on two-dimensional odometry.
  • step S301 The processing in step S301 will be described later with reference to FIGS.
  • the traveling locus correction unit 102 associates the movement coordinates of the overlapping traveling area, which is an area where the autonomous mobile device 10 travels in duplicate (S302).
  • the process of step S302 will be described later with reference to FIG.
  • the travel locus correction unit 102 performs the same process as in step S301 again using the result of step S302, and performs the second correction for correcting the travel locus (S303).
  • FIGS. 7 to 18 Details of the travel locus correction and map generation by the travel locus correction unit 102, the partial map generation unit 103, and the map generation unit 104 according to the present embodiment will be described.
  • FIGS. 7 to 18 show two-dimensional movement coordinates.
  • FIG. 7 is a diagram illustrating an example of a travel area of the autonomous mobile device according to the present embodiment.
  • the autonomous mobile system 1 acquires environmental information in the partial area 610 and maps a partial map obtained by mapping the partial area 610. Generate. As shown in FIG. 18, a plurality of partial maps are generated so that overlapping regions are generated in different partial regions 610, 901, and 902. And the autonomous mobile system 1 finally produces
  • the partial area 610 is an area where the autonomous mobile device 10 acquires environment information in one run. Reference numerals 611 to 613, 621, and 622 will be described later.
  • FIG. 8 is a diagram illustrating an example of a travel locus according to the present embodiment.
  • the autonomous mobile device 10 has moved along the route 701 in a single travel (for example, manual travel).
  • the true movement coordinates acquired by this traveling should be the movement coordinates on the route 701. That is, the travel locus should be the shape of the route 701.
  • a traveling locus based on the movement coordinates before correction is indicated by reference numeral 711.
  • errors are accumulated in the movement coordinates acquired by the environment information acquisition unit 101 due to the effects of slippage between the movement mechanism and the road surface, measurement errors of the sensor, and the like.
  • the travel locus 711 based on the movement coordinates acquired by the environment information acquisition unit 101 is distorted in shape. That is, the travel locus 711 based on the movement coordinates acquired by the environment information acquisition unit 101 is deviated from the true route 701.
  • the traveling locus correction unit 102 first corrects the traveling locus 711 based on the uncorrected movement coordinates by the constraint based on the GPS positioning information (hereinafter, more specifically referred to as GPS coordinates). .
  • the constraint based on the GPS coordinate indicates where the coordinate of each point of the movement coordinate corresponds to in an arbitrary absolute coordinate system (for example, a world coordinate system such as latitude, longitude, and altitude).
  • the travel locus correction unit 102 corrects the travel coordinates by deforming the travel locus 711 based on the travel coordinates before correction and acquiring the shape of the travel locus that minimizes the deviation from the GPS coordinates.
  • the travel locus correction unit 102 defines, for example, the error as a density function based on errors that can occur based on the physical meaning of each constraint. Then, the traveling locus correction unit 102 may consider that all constraints are satisfied to the maximum by changing the positional relationship of the movement coordinates so that the value based on the density function is maximized or minimized. Good. That is, the travel locus correction unit 102 considers that all constraints are satisfied to the maximum by the optimization process.
  • a method of solving as an optimization problem of simultaneous equations such as a Graph-SLAM (Simultaneous Location And Mapping) method may be used, or a random relationship with respect to the traveling locus may be used.
  • a technique based on a procedure that satisfies constraints by superimposing minute corrections by trial and error may be used.
  • FIG. 9 to FIG. 11 a specific procedure for deforming the travel locus 711 before correction to obtain the corrected travel locus will be described in detail.
  • FIG. 9 is a diagram showing a detailed procedure of the first correction according to the present embodiment.
  • a travel locus 711 based on a movement coordinate 721 before correction of the travel locus (referred to as a movement coordinate 721 before correction as appropriate) is obtained.
  • GPS coordinates are acquired as reference numeral 731.
  • “ ⁇ ” marks indicate movement coordinates measured by odometry
  • “x” marks indicate GPS coordinates (GPS coordinates).
  • the movement coordinates 721 before correction and the GPS coordinates 731 are associated with each other (for example, movement coordinates 721a and GPS coordinates 731a before correction, movement coordinates 721b and GPS coordinates 731b before correction, and the like).
  • the environment information acquisition unit 101 acquires information (movement coordinates) for generating a traveling locus of the autonomous mobile device and absolute coordinates (GPS coordinates) associated with a predetermined position of the traveling locus.
  • the traveling locus correction unit 102 obtains a corrected traveling locus indicated by a broken line by deforming the traveling locus 711 based on the movement coordinates 721 before correction and the GPS coordinates 731. For example, the travel locus correction unit 102 generates a corrected travel locus 741 that minimizes an error evaluation value defined from the movement coordinates 721 and the GPS coordinates 731 before correction. As the travel locus 711 is corrected, the movement coordinates 721 on the travel locus 711 are also corrected.
  • FIG. 10 is a diagram illustrating an example of an error evaluation value (first error evaluation value) used for the first correction according to the present embodiment.
  • FIG. 10 is an enlarged view of a part of the travel locus shown in FIG. 9, and the same components as those in FIG. 9 are denoted by the same reference numerals.
  • a point where the GPS coordinate 731 is not associated with the uncorrected movement coordinate 721 (hereinafter referred to as the movement coordinate 721 as appropriate) is a point where the GPS coordinate could not be acquired.
  • the travel locus correction unit 102 obtains the existence probability of a GPS coordinate defined by a density function with respect to the distance from the acquired GPS coordinate 731, and an area indicated by a value (2 ⁇ ) obtained by multiplying the standard deviation ( ⁇ ) by 2 (2 ⁇ region 732) is calculated for each GPS coordinate 731. Then, the travel locus correction unit 102 draws a line segment 733 from the point closest to the movement coordinate 721 in the 2 ⁇ region 732 to the movement coordinate 721. The travel locus correction unit 102 calculates the line segment 733 for each pair of movement coordinates 721 and GPS coordinates 731. At this time, the length of the line segment 733 is set to “0” at a point where the GPS coordinate 731 cannot be acquired.
  • the traveling locus correction unit 102 sets the square sum of each line segment 733 as an error evaluation value, and deforms the traveling locus 711 before correction (referred to as traveling locus 711 as appropriate) so that the square sum becomes the smallest. By doing so, a corrected travel locus 741 (FIG. 9) is obtained. At this time, it is preferable that the travel locus is deformed so that the travel locus 711 is as linear as possible except for a corner (for example, a portion where each of the travel locus 711 has a certain size or more).
  • the traveling locus correction unit 102 slightly deforms the traveling locus 711 to obtain a total value of the line segment 733 as an error evaluation value, and then deforms the traveling locus 711 further slightly to obtain the total value of the line segment 733. Is repeated to obtain a corrected travel locus 741 in which the total value of the line segment 733 is the smallest.
  • the reason for using such a method is that the GPS coordinate 731 also includes an error.
  • RTK-GPS Real Time Kinematic-GPS
  • D-GPS Downlink-GPS
  • RTK-GPS Real Time Kinematic-GPS
  • D-GPS Downlink-GPS
  • RTK-GPS Real Time Kinematic-GPS
  • D-GPS can be re-positioned while the autonomous mobile device 10 is moving, and can be easily used because the number of necessary satellites is small.
  • the height of the moving coordinate may be corrected by using the height information of the partial area 610 (FIG. 7) measured in advance with an aircraft or the like together with the GPS coordinates. The method will be described with reference to FIG.
  • FIG. 11 is a diagram illustrating an example of an error evaluation value used for correction in consideration of the height information of the traveling path acquired from the aircraft (hereinafter referred to as aircraft information 751) in addition to the GPS coordinates 731.
  • aircraft information 751 is a diagram illustrating an example of an error evaluation value used for correction in consideration of the height information of the traveling path acquired from the aircraft (hereinafter referred to as aircraft information 751) in addition to the GPS coordinates 731.
  • aircraft information 751 is a diagram illustrating an example of an error evaluation value used for correction in consideration of the height information of the traveling path acquired from the aircraft (hereinafter referred to as aircraft information 751) in addition to the GPS coordinates 731.
  • aircraft information 751 is a diagram illustrating an example of an error evaluation value used for correction in consideration of the height information of the traveling path acquired from the aircraft (hereinafter referred to as aircraft information 751) in addition to the GPS coordinates 731.
  • aircraft information 751 is a diagram illustrating an example of an error evaluation value used for correction in consideration of
  • the travel locus correction unit 102 calculates a 2 ⁇ region 752 for the aircraft information 751 in the same manner as the GPS coordinate 731 in FIG. 10, and the uncorrected movement coordinate 721 (appropriately, the movement coordinate 721 in the 2 ⁇ region 752).
  • a line segment 753 is drawn on the movement coordinate 721 from the point closest to the above.
  • the travel locus correction unit 102 calculates the line segment 753 for each pair of movement coordinates 721 and aircraft information 751. At this time, the length of the line segment 753 is “0” at a point where the aircraft coordinates 751 cannot be acquired.
  • the traveling locus correction unit 102 sets the total value of all the line segments 733 and all the line segments 753 as an error evaluation value (first error evaluation value), and the traveling locus before correction so that the total value becomes the smallest. 711 (referred to as travel locus 711 as appropriate) is modified. Thereby, the traveling locus correction unit 102 obtains a corrected traveling locus 741 (FIG. 9).
  • the travel locus may be deformed so that the travel locus 711 is as linear as possible except at a corner (for example, each portion formed by the travel locus 711 has a certain size or more).
  • the traveling locus correction unit 102 may add weight according to the accuracy of the GPS coordinates 731 and the aircraft information 751. For example, if the aircraft information 751 is more accurate than the GPS coordinate 731, the traveling locus correction unit 102 increases the weight of the line segment 753.
  • the correction of the travel locus shown in FIGS. 10 and 11 is an example, and other methods may be used.
  • the corrected travel locus 741 may be obtained by using a genetic algorithm, a neural network, or the like. The above is the travel locus correction processing in step S301 in FIG.
  • the movement coordinates are sequentially corrected as in the prior art, the movement coordinates are not corrected at locations where GPS coordinates are not acquired, such as tunnels, and errors are accumulated during that time. And since GPS coordinates can be acquired, the correction is made suddenly from the state in which the errors up to that point are accumulated, so that the travel locus becomes discontinuous.
  • the GPS coordinates are acquired from a place where the GPS coordinates are not acquired in order to correct the travel locus in consideration of the influence of other places. A continuous running track can be obtained even when the vehicle moves to a place.
  • the travel locus correction unit 102 calculates the first error evaluation value based on the absolute coordinates, and deforms the travel locus of the autonomous mobile device obtained from the relative coordinates based on the first error evaluation value. Thus, the first correction for correcting the travel locus is performed.
  • FIG. 12 is a diagram for explaining the detailed procedure of associating the movement coordinates of the overlapping travel area (step S302 in FIG. 5).
  • the travel trajectory correction unit 102 causes the autonomous mobile device 10 to pass through the vicinity of the corrected travel trajectory 741 generated in step S301 as shown in FIG. It is possible to obtain a region (overlapping traveling region 771) that can be estimated as being. That is, the traveling locus correction unit 102 detects the overlapping traveling region 771 by detecting a region where the corrected traveling locus is close.
  • the travel locus correction unit 102 has detected that the regions that are close to each other so that the tracks constituting the travel locus that has been corrected by the first correction are located, so that the autonomous mobile devices have traveled in duplicate. Detect overlapping running areas that are areas
  • the traveling locus correction unit 102 calculates the relative positional relationship between the movement coordinates 761 and 762, for example, by comparing the surrounding shape data obtained at the respective points of the movement coordinates 761 and 762.
  • shape data is shape data, such as a building acquired with a laser distance sensor, for example.
  • the shape data is information included in the environment information.
  • the travel locus correction unit 102 obtains the shape data (the thick line portion 765 in FIG. 12B) acquired by the autonomous mobile device 10 at the movement coordinates 761, and FIG. The shape data (the thick line portion 766 in FIG. 12C) acquired by the autonomous mobile device 10 at the movement coordinates 762 as shown in FIG.
  • the movement coordinates in the overlapping traveling area 771 are associated with each other, and the restriction between the moving coordinates in the overlapping traveling area 771 is determined.
  • the traveling locus correction unit 102 selects a plurality of relative coordinates on the traveling locus in the overlapping traveling region, and compares the environment information associated with each selected relative coordinate with each other to select each of the selected relative coordinates.
  • a first relative positional relationship between relative coordinates is calculated. Specifically, it is calculated by the processing in step S302 that the relative positional relationship between the movement coordinates 761 and 762 is a positional relationship as shown in FIG. 12D (broken arrows in FIG. 12D).
  • the constraint (relative positional relationship) between the two movement coordinates associated with each other in the overlapping travel area 771 is that one movement coordinate (for example, reference numeral 761) and the other movement coordinate ( For example, it has a meaning as to what value the position / posture of the autonomous mobile device 10 at 762) should satisfy.
  • the travel locus correction unit 102 associates the shape or pattern in the environment information obtained from the environment information acquisition unit 101 by an ICP (Iterative Closest Point) method in point cloud processing or a template matching method in image processing. Correspondence between moving coordinates is performed.
  • ICP Intelligent Closest Point
  • the traveling locus correction unit 102 can correct the movement coordinates so as to satisfy the constraints of the relative coordinate system.
  • a variation error occurs in the constraint based on the GPS coordinates, it is possible to obtain moving coordinates that are not repeatedly arranged at a location where the information on the same point is shifted. That is, the error does not disappear completely even in the result of the travel locus correction performed in step S301 in FIG.
  • the movement coordinates corrected in this way include an error, the overlapping traveling point is not recognized as another point.
  • the travel locus correction unit 102 first performs rough correction of the travel locus by eliminating the accumulation of errors from the travel locus 711 (moving coordinates) using GPS coordinates and aircraft information. First, do it. By doing in this way, it becomes possible to restrict the maximum value of the error that the moving coordinate has to the extent of the GPS coordinate error. Furthermore, in the overlapping traveling area in one traveling, the error at each associated point can be suppressed to the extent of the GPS coordinate error. Therefore, the certainty at the time of producing
  • the traveling locus correction unit 102 performs the second correction using the GPS coordinates again using the movement coordinates in which the overlapping traveling regions 771 are associated (step S303 in FIG. 5). That is, the travel locus correction unit 102 calculates a second error evaluation value based on both the calculated first relative positional relationship and the absolute coordinates, and uses the travel locus as the second error evaluation value. By performing the deformation based on the second correction, the second correction for correcting the travel locus is performed. At this time, the travel locus 711 is corrected again by the same procedure as in step S301.
  • FIG. 13 is a diagram illustrating an example of an error evaluation value (second error evaluation value) used for the second correction according to the present embodiment.
  • the travel locus correction unit 102 draws a line segment 783 from the movement coordinate 781 associated with the movement coordinate 721 in step S302 to the movement coordinate 721. At this time, the length and direction of the line segment 783 are drawn so as to be a relative positional relationship between the movement coordinates 721 and 781 calculated in step S302.
  • the travel locus correction unit 102 calculates the line segment 783 for each movement coordinate 721.
  • the travel locus correction unit 102 uses the total value of all the line segments 733 and all the line segments 783 as an error evaluation value, and deforms the travel locus 741 after the first correction so that the total value is minimized. Then, the second correction is performed. At this time, it is desirable to increase the weight of the line segment 783 (indicated by a thick line in FIG. 13). Similarly to FIG. 10, the travel locus may be deformed so that the travel locus 711 is as linear as possible except at a corner (for example, each of the travel locus 711 is a certain size or more).
  • the traveling locus correction unit 102 may draw a line segment 783 between the 2 ⁇ region and the movement coordinate 781 and use the total value of all the line segments 733 and all the line segments 783 as an error evaluation value.
  • the movement coordinates associated with each other in a relative positional relationship are paired for each transfer coordinate, but association between a plurality of movement coordinates may be used.
  • a line segment having a length based on the relative positional relationship may be used as the error evaluation value.
  • the travel locus correction unit 102 does not use the result of step S301, but performs processing from the beginning (that is, uses the travel locus before correction instead of using the corrected travel locus). It is desirable.
  • trajectory used by this embodiment is made into three types of the movement information acquired by the autonomous mobile device 10, GPS coordinates, and the environment information by an aircraft, it is not restricted to these three types, There may be four or more types. Moreover, as long as it is an absolute coordinate, information other than the GPS coordinate or the environment information by the aircraft may be used. In addition, it is desirable to set the traveling route of the autonomous mobile device 10 in advance so that the overlapping traveling area is as large as possible.
  • FIG. 14 is a diagram illustrating a change in the travel locus associated with the first correction and the second correction according to the present embodiment.
  • a travel trajectory 711 in FIG. 14A is a travel trajectory before the first correction, and is a movement coordinate itself acquired from odometry.
  • 741 shown in FIG. 14B is obtained by performing the first correction shown in FIGS. 9 to 11 on the travel locus 711.
  • the travel locus 741 is closer to the true travel locus (the shape of the route 701 in FIG. 8) than the travel locus 711, but the influence of errors in absolute coordinates such as GPS coordinates remains. In other words, the error in the travel locus 741 is suppressed to the extent of an error in absolute coordinates such as GPS coordinates.
  • the travel locus correction unit 102 detects an overlapping travel region for the travel locus 741 after the first correction, and the second correction shown in FIGS. 12 and 13 is performed, so that FIG. A travel locus 791 shown in FIG.
  • the traveling locus 791 is a true traveling locus (the shape of the route 701 in FIG. 8) than the traveling locus 711 or 741 It is close to.
  • a traveling path 601 that is an object of autonomous movement includes a region 612 in which a structure 611 such as a tree or a building exists, and a region 613 in which there is no structure around. .
  • balloons 621 and 622 from the areas 612 and 613 in FIG. 7 show bird's-eye views of the areas 612 and 613, respectively.
  • the environment information acquired by the aircraft is referred to as the sky environment information.
  • the aircraft information described above (the travel path height information acquired from the aircraft) is information included in the sky environment information.
  • FIG. 15 and FIG. 16 are diagrams showing conceptual diagrams for determining whether or not to use environment information acquired from the sky.
  • object shape data is used as environment information.
  • the aircraft information described above uses this shape data as height information.
  • FIG. 15 is a diagram related to a region where a structure 611 such as a tree or a building exists in the vicinity of the autonomous mobile environment (the region 612 in FIG. 7), and FIG.
  • FIG. 16 illustrates a structure such as a tree or a building around the autonomous mobile environment.
  • FIG. 8 is a diagram relating to a region in which there is no such as region 613 in FIG.
  • the environmental information is acquired from the aircraft 800 that acquires environmental information from the sky and the autonomous mobile device 10 that acquires environmental information while traveling on the ground. Is done.
  • the environment information obtained from the autonomous mobile device 10 is referred to as ground environment information.
  • a shape 821 shown in FIGS. 15A and 15B is a cross section of shape data obtained from the sky environment information in the region of FIG.
  • a shape 822 in FIGS. 15A and 15C is a cross section of shape data obtained from the ground environment information in the region of FIG.
  • a shape 831 shown in FIGS. 16A and 16B is a cross section of shape data obtained from the sky environment information in the region of FIG.
  • a shape 832 in FIGS. 16A and 16C is a cross section of shape data obtained from the ground environment information in the region of FIG.
  • shape data (shape) obtained from the sky environment information 821) and the shape data (shape 822) obtained from the ground environment information are different shapes.
  • the shape of the shape data (shape 832) obtained from the ground environment information is similar to (shape 831).
  • the shape data obtained from the sky environment information differs from the shape data obtained from the ground environment information depending on whether or not a structure exists in the surroundings.
  • the traveling locus correction unit 102 associates a shape / pattern or the like between the shape data obtained from the sky environment information and the shape data obtained from the ground environment information, and determines whether or not they are similar. To do. In this way, the travel locus correction unit 102 matches the height information in the movement coordinates acquired by the autonomous mobile device 10 with the height information acquired from the aircraft 800 or the like in the sky (the degree of coincidence is a predetermined value or more). Whether or not to do so can be determined. That is, the traveling locus correction unit 102 can determine whether or not the height in the traveling locus (moving coordinates) can be corrected based on the height information obtained from the sky environment information.
  • the traveling locus correction unit 102 The height information in the travel locus (movement coordinates) is not corrected with the height information based on the information.
  • the shape data (shape 831) based on the sky environment information is similar to the shape data (shape 832) obtained from the ground environment information.
  • the travel locus is corrected with the height information based on the environmental information.
  • the traveling locus correction unit 102 compares the shape information acquired by the aircraft with the shape information acquired by the autonomous mobile device itself, and the shape information acquired by the aircraft and the shape acquired by the autonomous mobile device itself. When the information matches, it is possible to correct the travel locus using the height information acquired by the aircraft at that point. This process is performed before step S301 in FIG.
  • the method for associating the shape data based on the ground environment information and the shape data based on the sky environment information used here may be any method.
  • the traveling locus correction unit 102 may calculate the sky environment information and the ground environment information. You may calculate a similarity directly between shape data.
  • the travel locus correction unit 102 may separate the road surface and the wall surface (three-dimensional object) from the ground environment information, and determine from the density of the wall surface existing in the surroundings and the similarity of the positions.
  • the traveling locus correction unit 102 may use the degree of similarity of the road surface pattern.
  • the edge (height change) of the shape data and the edge (luminance change) of the image may be associated with each other.
  • the constraint may be the absolute coordinate system, and the coordinate system used in the sky environment information and the terrestrial rights cooperative information
  • the coordinate system used in is different from that of the coordinate system, the constraint of both coordinate systems is used for correction as the constraint of the relative coordinate system.
  • the processes in FIGS. 15 and 16 can be omitted.
  • the travel locus including the correction of the height information can be corrected. Further, by matching the shape data acquired by the autonomous mobile device 10 with the shape data acquired from an aircraft or the like, and determining whether or not the height information acquired from the aircraft can be used, the accuracy in correcting the travel locus can be improved. Can be improved. Note that by using shape data acquired from a device flying over the aircraft 800 or the like, shape data of a wide area can be acquired at a time.
  • the partial map generation unit 103 adds shape data acquired from a laser distance sensor or the like as necessary based on the corrected movement coordinates of the partial region, and generates a partial map (see FIG. 3 step S103). That is, the partial map generation unit 103 generates a map based on the travel locus corrected by performing the second correction. And this map (partial map) is produced
  • the travel locus correction unit 102 can correct the three-dimensional travel locus acquired in one run and generate a partial map based on the corrected travel locus.
  • the movement coordinates in each partial map have no constraint on each other. That is, it is in a state in which it is not known what positional relationship the moving coordinates between the partial maps are.
  • the positioning error tends to be different because the arrangement of GPS satellites is greatly different.
  • the movement coordinate constraint between the partial maps is specified by the method shown in FIGS. 17 and 18.
  • FIG. 17 is a flowchart showing a partial map connection procedure according to this embodiment. This process shows details of step S104 in FIG. Details of each process will be described later with reference to FIG.
  • the map generation unit 104 associates movement coordinates in an overlapping area between maps that are overlapping areas between partial maps (S401). That is, each partial map is set with an inter-map overlapping area that is an overlapping area between adjacent maps.
  • the map generation unit 104 performs a third correction for correcting the travel locus in each partial map to be processed based on the movement coordinates associated in step S401 (S402). Subsequently, the map generation unit 104 combines the partial maps based on the movement coordinates on the travel locus corrected in step S402 (S403).
  • FIG. 18 is a diagram for explaining a partial map connection method according to the present embodiment.
  • a plurality of partial maps (partial regions 610, 901, 902) obtained by dividing the region where the autonomous mobile device travels are acquired so that an inter-map overlapping region that is an overlapping region between the partial maps is generated.
  • a partial map (partial region) is acquired so that the entire travel path 601 is covered.
  • the maximum value of the error is already limited for the movement coordinates in each of the partial areas 610, 901, and 902. Therefore, it is possible to obtain constraints between the moving coordinates in the relative coordinate system based on the above-described correspondence between the overlapping portions.
  • the map generation unit 104 further corrects the travel trajectory so as to satisfy the constraint of the relative coordinate system between the partial maps, and the moving coordinates that do not shift the connection points in the partial areas 610, 901, and 902 are obtained. obtain.
  • the movement coordinates 1111 exist on the corrected travel locus 1110 in the partial area 610 and the movement coordinates 1121 and 1122 exist on the corrected travel locus 1120 in the partial area 901.
  • the map generation unit 104 calculates a relative positional relationship (second relative positional relationship, that is, a constraint relationship) of the movement coordinates 1111, 1121, and 1122 from shape data such as surrounding buildings.
  • the map generation unit 104 compares the environment information in the inter-map overlap area with each other for adjacent maps, thereby obtaining a second relative position that is a relative positional relationship between the relative coordinates in the inter-map overlap area. Calculate the positional relationship.
  • the map generation unit 104 associates the movement coordinates in the inter-map overlap area using the same method as in step S302 in FIG.
  • the association of the positional relationship of the movement coordinates corresponds to the process of step S401 in FIG.
  • the map generation unit 104 determines the positional relationship between different points on the travel locus in the partial map sharing the map overlap region by comparing the surrounding environment information in the map overlap region. To do. By doing so, it is possible to specify the constraint (positional relationship) between the coordinates in the inter-map overlap region, so that the partial maps can be accurately combined without human intervention.
  • the map generation unit 104 corrects the travel locus so that the positional relationship of the movement coordinates 1111, 1121, and 1122 on the map becomes the determined positional relationship (third correction).
  • the map generation unit 104 deforms one or both traveling trajectories in adjacent maps so that each relative coordinate in the inter-map overlap area has the calculated second relative positional relationship, thereby generating a map.
  • a third correction for correcting the travel locus at is performed. This correction of the travel locus corresponds to step S402 in FIG. That is, the map generation unit 104 performs a third correction for correcting the travel locus on the partial map that is the processing target so that the points on the partial map have the determined positional relationship.
  • the map generation unit 104 translates or rotates the travel trajectory so that the movement coordinates on the travel trajectories are in a determined positional relationship.
  • the map generation unit 104 may adjust the influence on the constraint according to the distance / distance to each moving coordinate. In other words, the correction may be performed in accordance with the amount of distance between the travel coordinates and the movement coordinates, that is, after adjusting by weighting the distance between the distance of travel and the movement coordinates.
  • the map generation unit 104 increases the weight of the positional relationship (constraint relationship) of the movement coordinates between the partial maps, and then performs the processing of steps S301 to S303 in FIG. 5 to correct the travel locus again. It may be. By doing in this way, the certainty in the connection of a partial map can be improved.
  • the certainty of the connection of the partial maps can be improved by further correcting the traveling trajectory so as to satisfy the constraint of the relative coordinate system between the partial maps.
  • the map generation unit 104 then combines the partial maps based on the movement coordinates on the travel locus corrected in step S402 (step S403 in FIG. 17). In other words, the map generation unit 104 combines the maps by combining the relative coordinates in the travel locus corrected by the third correction in the inter-map overlap area.
  • the map generation unit 104 completes a map for self-position estimation by arranging shape data such as buildings around each moving coordinate obtained as necessary.
  • the completed map includes road surface height information (height information) at each point. Note that, even if they are at the same point, if the heights of the respective movement coordinates do not completely match due to errors or the like, the map generation unit 104 calculates an average value from adjacent values and calculates the height at an arbitrary point. You may consider that.
  • the movement coordinates when moving autonomously are corrected by the above-described method, and the corrected driving is performed.
  • the environment is acquired by traveling many times on the travel path 601 (FIG. 7) that is the target of autonomous movement, constraints are increased, so that variation in each error can be reduced by averaging. By doing in this way, a map can be highly accurate.
  • the travel locus correction unit 102 determines the constraints between the movement coordinates obtained by the autonomous mobile device 10 traveling the same place many times during the autonomous movement, using the processing of steps S301 to S303 in FIG. Thus, it is possible to obtain a corrected travel locus with higher accuracy.
  • environment information such as moving coordinates increases too much, it may be reduced by thinning or compression, and information with low existence probability when acquiring environment information multiple times is assumed to be noise such as moving objects You may make it delete preferentially.
  • working locus is performed based on a movement coordinate and a GPS coordinate
  • environmental information by GPS and aircraft is used to correct the height information in the travel locus.
  • the present invention is not limited to this, and environmental information measured by other autonomous mobile devices 10 may be used. Good.
  • the travel locus may not be represented by a line as in the present embodiment.
  • the travel locus may be displayed as a point sequence.
  • Each of the above-described configurations, functions, units 101 to 107, storage units 121 and 122, etc. may be realized by hardware by designing a part or all of them, for example, with an integrated circuit. Further, as shown in FIG. 2, the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by a processor such as the CPU 201, 211. Information such as programs, tables, and files for realizing each function is stored in the memory 202, the ROM 213, and the HD 215 as shown in FIG.
  • control lines and information lines are those that are considered necessary for explanation, and not all control lines and information lines are necessarily shown on the product. In practice, it can be considered that almost all configurations are connected to each other.
  • the error of the movement coordinate can be suppressed to the extent of the error of the GPS coordinate, so that the certainty of the partial map connection can be improved. Further, by performing the process of step S301 in FIG. 5 and correcting the overall travel locus, even if there is a point where GPS coordinates or aircraft information cannot be obtained, the originally continuous travel locus is discontinuous. Can be prevented.
  • step S301 in FIG. 5 of the present embodiment makes it possible to detect the overlapping traveling area, and it is possible to specify the constraint (relative positional relationship) between the movement coordinates in the overlapping traveling area in step S302.
  • the second correction is performed after associating the movement coordinates, so that the error of the movement coordinates can be made more accurate. It can be corrected.
  • steps S301 to S303 in FIG. 5 it is possible to associate the movement coordinates in the inter-map overlap area between the partial maps without the user's hand.
  • the travel locus corrected in this embodiment By generating a partial map using the travel locus corrected in this embodiment and combining the partial maps, it is possible to combine with high accuracy and generate a highly accurate wide area map. .
  • the height information in the movement coordinates By including the height information in the movement coordinates, it is possible to correct the travel locus including the height direction.
  • the movement coordinates are based on dead reckoning or odometry, the travel locus from the viewpoint of the autonomous mobile device 10 can be corrected.
  • realization of this embodiment can be made easy by using general latitude and longitude for an absolute coordinate system.
  • height information can be easily obtained by using environmental information obtained from an aircraft as height information.
  • the accuracy in correcting the height information is determined by matching the shape data acquired by the autonomous mobile device 10 with the shape data acquired from the aircraft or the like to determine whether or not the height information acquired from the aircraft can be used. Can be improved.

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
  • Instructional Devices (AREA)

Abstract

The present invention addresses the problem of improving reliability in connections of partial maps. The present invention is characterized in having: an environment information acquisition unit (101) which associates and acquires relative coordinates, absolute coordinates, and ambient environment information; and a traveling trajectory correction unit (102). The traveling trajectory correction unit (102): calculates a first error evaluation value that is based on the absolute coordinates; performs a first correction which, by deforming the traveling trajectory on the basis of the first error evaluation value, corrects the traveling trajectory; detects an overlapping travel area in which traveling trajectories that have been corrected by the first correction approach each other; by mutually comparing environment information units together in the overlapping travel area, calculates relative positional relationships between each of the selected relative coordinates; calculates a second error evaluation value that is based on the relative positional relationships and the absolute coordinates; and by deforming the traveling trajectory on the basis of the second error evaluation value, performs second correction which corrects the traveling trajectory.

Description

自律移動システムおよび管制装置Autonomous mobile system and control device
　本発明は、環境情報と地図を基に、自己位置を推定し、推定した自己位置を基に自律移動する自律移動システムおよび管制装置の技術に関する。 The present invention relates to a technology of an autonomous movement system and a control device that estimates a self-position based on environmental information and a map and autonomously moves based on the estimated self-position.
　人が操縦しなくともある地点から別の地点まで自動的に移動可能な自律移動装置は、工場内、建設現場、鉱山などの物品搬送や街中における次世代交通として、安全性・利便性・効率性の向上の観点から導入が望まれている。 Autonomous mobile devices that can move automatically from one point to another without maneuvering are safe, convenient, and efficient as next-generation transportation in factories, construction sites, and mines, and in cities. Introduction is desired from the viewpoint of improving the performance.
　自律移動装置の実現には、自律移動装置自身が、どの位置に存在するかを正確に把握して走行位置を間違えないための自己位置推定技術を備えたシステムが必要である。自己位置推定技術では、予め用意された、例えば画像情報や、形状情報や、座標などの環境情報が自己位置推定用の地図（以下、単に地図とも称する）として用いられる。自律移動装置は、自律移動中に自律移動装置自身の周囲から取得した環境情報と、予め用意されている地図とを照らし合わせることで地図上における自律移動装置自身の位置を特定する。このため、地図を高精度に生成することで自己位置推定精度が向上し、自律移動装置は安定した自律移動を行うことができる。ただし、予め用意された地図と、自律移動中に周囲から取得された環境情報との照合の確実性を向上させるため、地図が持つ情報は自律移動装置から取得したものであることが望ましい。 Realization of an autonomous mobile device requires a system equipped with a self-position estimation technique for accurately grasping where the autonomous mobile device itself is located and not making a mistake in the traveling position. In the self-position estimation technique, environmental information such as image information, shape information, and coordinates prepared in advance is used as a map for self-position estimation (hereinafter also simply referred to as a map). The autonomous mobile device identifies the position of the autonomous mobile device itself on the map by comparing environmental information acquired from the surroundings of the autonomous mobile device itself with a map prepared in advance. For this reason, self-position estimation accuracy improves by generating a map with high accuracy, and the autonomous mobile device can perform stable autonomous movement. However, in order to improve the certainty of matching between a map prepared in advance and environmental information acquired from the surroundings during autonomous movement, it is desirable that the information held in the map is acquired from the autonomous mobile device.
　例えば、衛星や航空機を用いて上空から取得した環境情報を基に生成された地図は、地上（床面あるいは路面）を移動する自律移動装置とは環境の見え方が大きく異なるため、そのままでは自律移動のための地図として利用できない。また同様に、同じ高さであっても遠く離れた地点から取得した環境情報を基に生成した地図は、やはり、自律移動装置とは見え方が異なる可能性が高いため地図として利用できない。このため、自律移動の対象とする環境で、予め自律移動装置を手動走行（人による操縦）させて、周囲の環境情報を取得し、この環境情報を基に地図を生成しておくことが一般的に行われている。このようにすることで、自律移動時の照合において確実性の高い地図の生成が可能となる。 For example, maps generated based on environmental information acquired from the sky using satellites and aircraft have a much different environment appearance than autonomous mobile devices that move on the ground (floor surface or road surface). It cannot be used as a map for moving. Similarly, a map generated based on environmental information acquired from a distant point even at the same height is not likely to be used as a map because it is likely to look different from the autonomous mobile device. For this reason, it is common to manually run an autonomous mobile device (maneuvered by a person) in advance in an environment subject to autonomous movement, acquire surrounding environment information, and generate a map based on this environment information. Has been done. By doing in this way, it becomes possible to generate a map with high certainty in collation during autonomous movement.
　ただし、手動で自律移動装置を走行させて取得した環境情報から高精度な地図の生成する場合、地図の広域化において、次のような課題が存在する。それは、広域的な地図を生成する際に、地図生成対象範囲全域の環境情報を一度の手動走行で取得することが、精度や時間の観点から困難であるということである。このため、自律移動を行う範囲である地図生成対象範囲を複数回に分けて取得した部分地図を正確に接続・統合することで、広域的な地図の生成が行われる。 However, when a highly accurate map is generated from environmental information acquired by running an autonomous mobile device manually, the following problems exist in widening the map. That is, when generating a wide-area map, it is difficult from the viewpoint of accuracy and time to acquire environmental information of the entire map generation target range by one manual travel. For this reason, a wide-area map is generated by accurately connecting and integrating the partial maps acquired by dividing the map generation target range, which is a range where autonomous movement is performed, into multiple times.
　このような地図の生成方法として、複数に分割して取得した地図を正確に接続・統合するための地図生成方法が開示されている（例えば、特許文献１および特許文献２参照）。特許文献１に記載の技術は、環境から１つ以上の特徴点を抽出し、分割された地図間で特徴点を対応付けることで拘束関係を導き、これに基づいて１つの広域的な地図を生成している。 As such a map generation method, a map generation method for accurately connecting and integrating maps obtained by dividing into a plurality of maps is disclosed (for example, see Patent Document 1 and Patent Document 2). The technique described in Patent Document 1 extracts one or more feature points from the environment, derives a constraint relationship by associating the feature points between the divided maps, and generates one wide-area map based on this. is doing.
　具体的には、特許文献１に記載の技術は、自律移動する環境を重複のない部分地図として分割して取得し、各々の特徴情報の拘束関係に基づいて相互の位置関係を決定して補正することで１つの地図を生成する。部分地図同士は交差点や曲がり角など環境形状が特徴的な狭い領域毎に設定される。拘束関係は隣接する部分地図の位置関係や形状の平行性などをユーザが入力するか、周辺環境の設計図に基づいてユーザが設定することで地図を生成する構成としている。 Specifically, the technology described in Patent Document 1 acquires and divides an autonomously moving environment as a partial map without duplication, and determines and corrects the mutual positional relationship based on the constraint relationship of each feature information. By doing so, one map is generated. The partial maps are set for each narrow area having a characteristic environmental shape such as an intersection or a corner. The constraint relationship is configured such that the map is generated by the user inputting the positional relationship of the adjacent partial maps, the parallelism of the shape, or the like based on the design drawing of the surrounding environment.
ここで、拘束関係とは、例えば、同じ地点が含まれている異なる環境情報間で、その地点を結びつけるため、その複数の情報間で位置が拘束される関係をいう。また、同じ地点が含まれている異なる環境情報とは、例えば、異なる部分地図の同じ地点の環境情報や、ひとつもしくは複数の部分地図で異なる経路や手段によって取得された同じ地点の環境情報なども拘束関係と称する。 Here, the constraint relationship refers to a relationship in which positions are constrained between a plurality of pieces of information in order to link the locations between different environmental information including the same location. In addition, different environmental information including the same point includes, for example, environmental information of the same point in different partial maps, environmental information of the same point acquired by different routes and means in one or more partial maps, etc. This is called a restraint relationship.
　具体的には、拘束は、２つ以上の情報間の位置・姿勢などの関係を意味し、例えば、２つ（あるいは複数）の重複している地図間において、一方の地図上の座標が、他方の地図上のどの座標に対し、どのような位置関係にあるのかを示すものである。あるいは、例えば、ＧＰＳ（Global Positioning System）測位情報とオドメトリなどの移動座標の間における拘束は、ある移動座標がＧＰＳ測位情報における絶対座標系においてどの地点であるかというものも拘束関係と称する。 Specifically, the constraint means a relationship such as position / posture between two or more pieces of information. For example, between two (or a plurality of) overlapping maps, the coordinates on one map are It shows what kind of positional relationship it is with respect to which coordinate on the other map. Alternatively, for example, a constraint between GPS (Global Positioning System) positioning information and movement coordinates such as odometry is also referred to as a constraint relationship as to which point a certain movement coordinate is in the absolute coordinate system in the GPS positioning information.
　また、特許文献２に記載の技術は、自律移動装置が備えている画面出力や、ボタン入力で人間が操作するインタフェースにより、複数の地図間の連結点を対応付けることで、１つの広域的な地図を生成している。 In addition, the technique described in Patent Document 2 is based on the screen output provided in the autonomous mobile device and the interface operated by a human by button input, thereby associating connection points between a plurality of maps into one wide area map. Is generated.
　具体的には、特許文献２に記載の技術は、特許文献１に記載の技術と同様に自律移動する環境を部分地図として分割して取得する。この部分地図間は僅かずつ重複し、画面出力やボタン入力により隣接する部分地図の位置や向きの拘束関係をユーザが設定することで１つの広域的な地図を生成する。特許文献２に記載の技術において、自律移動装置は、部分地図を乗り換えて目的とする地点へ移動する構成となっている。また、特許文献２に記載の技術は、環境の形状や走行ルートに変更がある場合には、その地点が含まれる部分地図のみを取得しなおして地図を再度生成する。 Specifically, the technology described in Patent Document 2 acquires an environment in which autonomous movement is performed as a partial map in the same manner as the technology described in Patent Document 1. The partial maps overlap each other little by little, and one wide-area map is generated by the user setting the constraint relationship between the positions and orientations of adjacent partial maps by screen output or button input. In the technique described in Patent Document 2, the autonomous mobile device is configured to change partial maps and move to a target point. In addition, in the technique described in Patent Document 2, when there is a change in the shape of the environment or the travel route, only the partial map including the point is acquired again and the map is generated again.
特開２００９－５３５６１号公報JP 2009-53561 A 特開２０１０－９２１４７号公報JP 2010-92147 A
　特許文献１に記載の技術は、複雑な形状の構造物がある場合、直線部分が少なくなるので平行性などの特徴情報を抽出しづらくなり、特徴情報の利用が困難となる。その結果、特許文献１に記載の技術は、部分地図を正確に接続することができないという状況が発生する。そのため、特許文献１に記載の技術は、複雑な形状の構造物がある場合において、広域的な地図作成には適用できない。また、特許文献１に記載の技術は、環境の路面が水平であることを前提としており、斜面や立体交差を有する場合については考慮されていない。そのため、特許文献１に記載の技術において、仮に環境情報が３次元情報を有している場合、生成される地図では高さ方向にずれ（誤差）が生じてしまう。 The technique described in Patent Document 1 makes it difficult to extract feature information such as parallelism because there are fewer straight portions when there is a complex-shaped structure, making it difficult to use the feature information. As a result, the technique described in Patent Document 1 causes a situation where the partial maps cannot be accurately connected. Therefore, the technique described in Patent Document 1 cannot be applied to wide-area map creation when there is a complex-shaped structure. In addition, the technique described in Patent Document 1 is based on the premise that the road surface of the environment is horizontal, and does not take into account the case of having a slope or a three-dimensional intersection. Therefore, in the technique described in Patent Document 1, if the environment information includes three-dimensional information, a shift (error) occurs in the height direction in the generated map.
　特許文献２に記載の技術では、１つ１つの部分地図に歪みが残されたままである。ここで、歪みとは、オドメトリなどの計測誤差に起因する誤差である。このように、特許文献２に記載の技術は、ライン状の走行軌跡であれば、単純な拘束関係を定義すればよいので、広域的な地図生成に対応できる。しかしながら、メッシュ状の走行軌跡に対応するには、複雑な拘束関係を定義する必要があるため、特許文献２に記載の技術は、部分地図をさらに細分化して多数の拘束関係を設定する必要がある。 In the technique described in Patent Document 2, distortion remains in each partial map. Here, the distortion is an error caused by a measurement error such as odometry. As described above, the technique described in Patent Document 2 can be used for a wide-area map generation because a simple constraint relationship may be defined as long as it is a line-like traveling locus. However, since it is necessary to define a complicated constraint relationship in order to cope with a mesh-like traveling locus, the technique described in Patent Document 2 needs to further subdivide the partial map and set a large number of constraint relationships. is there.
　このため、特許文献２に記載の技術は、メッシュ状の走行軌跡に対応するための広域な地図生成に適用できない。また、特許文献２に記載の技術は、特許文献１に記載の技術と同様に環境の路面が水平であることを前提としており、斜面や立体交差など環境情報が３次元情報を有する場合については考慮されていない。 For this reason, the technique described in Patent Document 2 cannot be applied to wide-area map generation for dealing with a mesh-like traveling locus. Moreover, the technique described in Patent Document 2 is based on the premise that the road surface of the environment is horizontal as in the technique described in Patent Document 1, and the case where the environment information such as slopes and three-dimensional intersections has three-dimensional information. Not considered.
　以上より、各々の部分地図における歪みを３次元的に除去することで実際の環境を正確に表現し、さらに複数の部分地図を統合した際に重複する地点が正確に接続されて連続性を有するようにすることで、高精度な地図の生成を行うことが必要である。 From the above, the actual environment is accurately expressed by removing the distortion in each partial map three-dimensionally, and when multiple partial maps are integrated, overlapping points are accurately connected and have continuity By doing so, it is necessary to generate a highly accurate map.
　このような背景に鑑みて本発明がなされたのであり、本発明は、部分地図の接続における確実性を向上させることを課題とする。 The present invention has been made in view of such a background, and an object of the present invention is to improve the certainty in connection of partial maps.
　前記課題を解決するため、本発明は、絶対座標を基に、走行軌跡を変形させる第１の補正を行い、さらに、補正した走行軌跡同士が近接している重複走行領域を検出し、この重複走行領域における走行軌跡上の相対座標を対応付けた後、さらに走行軌跡の補正を行う第２の補正を行うことを特徴とする。 In order to solve the above-described problem, the present invention performs a first correction for deforming a travel locus based on absolute coordinates, and further detects an overlapping travel region in which the corrected travel tracks are close to each other. After associating the relative coordinates on the travel locus in the travel region, a second correction for further correcting the travel locus is performed.
　本発明によれば、部分地図の接続における確実性を向上させることができる。 According to the present invention, certainty in connection of partial maps can be improved.
本実施形態に係る自律移動システムの構成例を示す図である。It is a figure which shows the structural example of the autonomous mobile system which concerns on this embodiment. 本実施形態に係る自律移動装置および管制装置のハードウェア構成例を示す図である。It is a figure which shows the hardware structural example of the autonomous mobile device and control device which concern on this embodiment. 本実施形態に係る地図生成処理の手順を示すフローチャートである。It is a flowchart which shows the procedure of the map production | generation process which concerns on this embodiment. 本実施形態に係る自律移動処理の手順を示すフローチャートである。It is a flowchart which shows the procedure of the autonomous movement process which concerns on this embodiment. 本実施形態に係る走行軌跡補正処理の詳細な手順を示す図である。It is a figure which shows the detailed procedure of the driving | running | working locus | trajectory correction process which concerns on this embodiment. 本実施形態に係る移動座標による走行軌跡の例を示す図である。It is a figure which shows the example of the driving | running | working locus by the movement coordinate which concerns on this embodiment. 本実施形態に係る自律移動装置の走行領域の例を示す図である。It is a figure which shows the example of the driving | running | working area | region of the autonomous mobile device which concerns on this embodiment. 本実施形態に係る走行軌跡の例を示す図である。It is a figure which shows the example of the traveling locus which concerns on this embodiment. 本実施形態に係る第１の補正の詳細な手順を示す図である。It is a figure which shows the detailed procedure of the 1st correction | amendment which concerns on this embodiment. 本実施形態に係る第１の補正に用いられる誤差評価値の一例を示す図である。It is a figure which shows an example of the error evaluation value used for the 1st correction | amendment which concerns on this embodiment. 本実施形態に係る第１の補正に用いられる誤差評価値の他の一例を示す図である。It is a figure which shows another example of the error evaluation value used for the 1st correction | amendment which concerns on this embodiment. 重複走行領域の移動座標の対応付けの詳細な手順を説明する図である。It is a figure explaining the detailed procedure of matching of the movement coordinate of an overlap driving | running | working area | region. 本実施形態に係る第２の補正に用いられる誤差評価値の一例を示す図である。It is a figure which shows an example of the error evaluation value used for the 2nd correction | amendment which concerns on this embodiment. 本実施形態に係る第１の補正、第２の補正に伴う走行軌跡の変化を示す図である。It is a figure which shows the change of the driving | running | working locus accompanying the 1st correction | amendment which concerns on this embodiment, and a 2nd correction | amendment. 上空から取得した環境情報の利用可否判定の概念図を示す図（その１）である。FIG. 10 is a diagram (part 1) illustrating a conceptual diagram of determination of availability of environmental information acquired from the sky. 上空から取得した環境情報の利用可否判定の概念図を示す図（その２）である。It is FIG. (2) which shows the conceptual diagram of the availability determination of the environmental information acquired from the sky. 本実施形態に係る部分地図の接続の手順を示すフローチャートである。It is a flowchart which shows the procedure of the connection of the partial map which concerns on this embodiment. 本実施形態に係る部分地図の接続方法を説明するための図である。It is a figure for demonstrating the connection method of the partial map which concerns on this embodiment.
　次に、本発明を実施するための形態（「実施形態」という）について、適宜図面を参照しながら詳細に説明する。なお、各図面において、同様の構成要素については、同一の符号を付して説明を省略する。
　本実施形態に係る自律移動システムは、自律移動装置を手動走行させた際の移動座標（移動軌跡）を高さ方向も含めて補正することで、移動座標を正確に導出可能とするものである。また、本実施形態に係る自律移動システムは、自律移動の環境が広域であっても、分割して取得した形状情報（部分地図）を正確に接続・統合することで自己位置推定用の地図生成技術を可能とすることにより、高精度な自己位置推定技術を実現するものである。なお、本実施形態において、部分地図は、自律移動装置が走行する領域を分割した部分領域に関する地図であり、各部分地図間には重複領域（地図間重複領域）が生じるよう設定されている。地図間重複領域については、図１８を参照して後記する。 Next, modes for carrying out the present invention (referred to as “embodiments”) will be described in detail with reference to the drawings as appropriate. In addition, in each drawing, about the same component, the same code | symbol is attached | subjected and description is abbreviate | omitted.
The autonomous mobile system according to the present embodiment makes it possible to accurately derive the movement coordinates by correcting the movement coordinates (movement locus) when the autonomous movement apparatus is manually driven, including the height direction. . In addition, the autonomous mobile system according to this embodiment generates a map for self-position estimation by accurately connecting and integrating shape information (partial maps) obtained by dividing even if the environment of autonomous movement is wide area By enabling the technology, a highly accurate self-position estimation technology is realized. In the present embodiment, the partial map is a map related to a partial region obtained by dividing the region where the autonomous mobile device travels, and is set so that overlapping regions (inter-map overlapping regions) are generated between the partial maps. The inter-map overlap area will be described later with reference to FIG.
［システム構成］
　図１は、本実施形態に係る自律移動システムの構成例を示す図である。
　自律移動システム１は、環境情報取得部（環境情報取得手段）１０１、走行軌跡補正部（走行軌跡補正手段）１０２、部分地図生成部（地図生成手段）１０３、地図生成部（地図生成手段）１０４、自己位置推定部１０５、経路生成部１０６、移動制御部１０７、環境情報記憶部１２１、地図記憶部１２２を有する。 [System configuration]
FIG. 1 is a diagram illustrating a configuration example of an autonomous mobile system according to the present embodiment.
The autonomous mobile system 1 includes an environment information acquisition unit (environment information acquisition unit) 101, a travel locus correction unit (travel locus correction unit) 102, a partial map generation unit (map generation unit) 103, and a map generation unit (map generation unit) 104. A self-position estimation unit 105, a route generation unit 106, a movement control unit 107, an environment information storage unit 121, and a map storage unit 122.
　環境情報取得部１０１は、複数の計測部を組み合わせて構成されている。環境情報取得部１０１は、例えば、レーザ距離センサ、単眼あるいは複眼のカメラシステム、ＧＰＳセンサ、気圧センサなどを有している。また、航空機や衛星から取得した画像や、自律移動装置１０の外部に設けたレーザ測量装置（不図示）からレーザ測量情報を取得してもよい。あるいは、環境情報取得部１０１は、別途人手にて測量した情報を取得してもよい。さらに、環境情報取得部１０１には自律移動装置１０から外界の情報を取得する装置の他に、自律移動装置１０の内部の情報取得する装置を組み合わせてもよい。このような環境情報取得部１０１として、例えば、自律移動装置１０が移動するための車輪、クローラ、脚などの移動機構から取得可能な移動量情報や、角速度や加速度などの運動量情報を取得可能な慣性計測センサなどがある。車輪、クローラ、脚などの移動機構から取得可能な移動量情報や、角速度や加速度などの運動量情報を、移動座標と称する。 The environment information acquisition unit 101 is configured by combining a plurality of measurement units. The environment information acquisition unit 101 includes, for example, a laser distance sensor, a monocular or compound eye camera system, a GPS sensor, an atmospheric pressure sensor, and the like. Further, laser survey information may be acquired from an image acquired from an aircraft or a satellite, or a laser surveying device (not shown) provided outside the autonomous mobile device 10. Alternatively, the environment information acquisition unit 101 may acquire information manually surveyed separately. Furthermore, the environment information acquisition unit 101 may be combined with a device that acquires information inside the autonomous mobile device 10 in addition to a device that acquires external information from the autonomous mobile device 10. As such an environment information acquisition unit 101, for example, it is possible to acquire movement amount information that can be acquired from a moving mechanism such as a wheel, crawler, or leg for movement of the autonomous mobile device 10, and momentum information such as angular velocity or acceleration. There are inertial measurement sensors. Movement amount information that can be acquired from movement mechanisms such as wheels, crawlers, and legs, and momentum information such as angular velocity and acceleration are referred to as movement coordinates.
　環境情報取得部１０１が取得した各環境情報（ＧＰＳセンサによるＧＰＳ座標や、レーザ距離センサによる周囲の形状情報や、航空機から取得される形状情報等）は、移動座標に対応付けられて環境情報記憶部１２１に格納される。環境情報取得部１０１は、自律移動装置の移動中における自身の相対的な位置を示す相対座標と、移動中における自身の絶対的な位置を示す絶対座標と、移動中における自身の周囲の環境情報と、を対応づけて取得する Each environment information acquired by the environment information acquisition unit 101 (GPS coordinates by the GPS sensor, surrounding shape information by the laser distance sensor, shape information acquired from the aircraft, etc.) is associated with the movement coordinates and stored in the environment information. Stored in the unit 121. The environment information acquisition unit 101 includes a relative coordinate that indicates the relative position of the autonomous mobile device during movement, an absolute coordinate that indicates the absolute position of the autonomous mobile device during movement, and environmental information of the surrounding area during movement. And get
　ここで、環境情報記憶部１２１には、自律移動装置１０がユーザによって、操縦・走行された際に得られた環境情報（手動環境情報）も格納されているし、自律移動装置１０が自律移動する際に得られた環境情報（自律環境情報）も格納されている。 Here, the environment information storage unit 121 also stores environment information (manual environment information) obtained when the autonomous mobile device 10 is steered and traveled by the user, and the autonomous mobile device 10 moves autonomously. The environment information (autonomous environment information) obtained when doing so is also stored.
　これらの環境情報は地図生成時には地図の生成に用いられ、自律移動時には、後記する自己位置推定部１０５が自己位置の推定に用いたり、後記する経路生成部１０６が周囲の構造物の配置や路面状況を把握して適切な経路を移動したりする際に用いられる。 These environmental information is used for map generation at the time of map generation. During autonomous movement, the self-position estimation unit 105 described later is used for estimation of the self-position, or the path generation unit 106 described later is used for arrangement of surrounding structures and road surfaces. It is used when grasping the situation and moving an appropriate route.
　走行軌跡補正部１０２は、環境情報取得部１０１が取得した複数種類の環境情報を基に、走行軌跡（詳細は後記）の補正を行う。
　部分地図生成部１０３は、走行軌跡補正部１０２によって補正された走行軌跡を基に、環境情報に含まれる形状データを補完して部分地図を生成する。部分地図生成部１０３は、手動環境情報のみを使用して部分地図を生成してもよいし、手動環境情報および自律環境情報の両方を基に用いて部分地図を生成してもよい。また、部分地図生成部１０３は、図示しない自律移動装置１０の外部に設けられている環境情報を取得する装置から環境情報をダイレクトに用いて部分地図を生成してもよい。 The travel locus correction unit 102 corrects the travel locus (described later in detail) based on a plurality of types of environment information acquired by the environment information acquisition unit 101.
The partial map generation unit 103 generates a partial map by complementing the shape data included in the environment information based on the travel locus corrected by the travel locus correction unit 102. The partial map generation unit 103 may generate a partial map using only manual environment information, or may generate a partial map using both manual environment information and autonomous environment information. In addition, the partial map generation unit 103 may generate a partial map by directly using environmental information from a device that acquires environmental information provided outside the autonomous mobile device 10 (not shown).
　地図生成部１０４では、部分地図生成部１０３で生成された複数の部分地図を接続して自己位置推定用の地図を生成する。ここで、地図生成部１０４は、自律移動装置１０の移動時にオンラインで地図を生成・更新してもよいし、オフラインで地図を生成・更新してもよい。
　生成された地図は、地図記憶部１２２に格納される。 The map generation unit 104 connects a plurality of partial maps generated by the partial map generation unit 103 to generate a map for self-position estimation. Here, the map generation unit 104 may generate / update a map online when the autonomous mobile device 10 moves, or may generate / update a map offline.
The generated map is stored in the map storage unit 122.
　自己位置推定部１０５は、自律移動中に取得した環境情報と地図記憶部１２２に記憶されている地図との対応付け（マッチング）によって自己位置を推定する。自己位置推定の手法として、前記した手法以外にも、車体移動量やその他の内部情報を用いて情報の累積による自己位置推定が用いられてもよいし、ＧＰＳ測位情報に基づく自己位置推定が用いられてもよい。また、これらの各自己位置推定手法を組み合わせた上で、フィルタリング処理（例えば、カルマンフィルタリングやその応用手法）が適用されることによって、各手法による自己位置推定結果が融合されてもよい。なお、地図との対応付けでは、画像処理や点群処理に用いられる手法が用いられる。 The self-position estimation unit 105 estimates the self-position by matching (matching) the environment information acquired during autonomous movement with the map stored in the map storage unit 122. As a self-position estimation technique, in addition to the above-described technique, self-position estimation based on accumulation of information using the vehicle body movement amount or other internal information may be used, or self-position estimation based on GPS positioning information is used. May be. Further, by combining these self-position estimation methods and applying a filtering process (for example, Kalman filtering or an application method thereof), the self-position estimation results by the respective methods may be merged. In association with the map, techniques used for image processing and point cloud processing are used.
　経路生成部１０６は、自律移動中に取得した環境情報から周囲の障害物や、路面の段差や、歩行者の位置や、歩行者の移動速度を抽出し、グラフ探索処理や運動モデルシミュレーション処理を用いることによって、自律移動装置１０の移動方向および速度を決定する。
　移動制御部１０７は、経路生成部１０６によって決定された自律移動装置１０の移動方向および速度に基づいて、自律移動装置１０を移動させる。これにより、自律移動装置１０は、ある地点から目的とする別の地点へ自動的に移動する。 The route generation unit 106 extracts surrounding obstacles, road surface steps, pedestrian positions, and pedestrian movement speeds from environmental information acquired during autonomous movement, and performs graph search processing and motion model simulation processing. By using it, the moving direction and speed of the autonomous mobile device 10 are determined.
The movement control unit 107 moves the autonomous mobile device 10 based on the moving direction and speed of the autonomous mobile device 10 determined by the route generation unit 106. Thereby, the autonomous mobile device 10 automatically moves from one point to another target point.
　図１に示すように、環境情報取得部１０１、自己位置推定部１０５、経路生成部１０６、移動制御部１０７、環境情報記憶部１２１および地図記憶部１２２が自律移動装置１０に搭載されることが考えられる。同様に、環境情報記憶部１２１、走行軌跡補正部１０２、部分地図生成部１０３、地図生成部１０４、地図記憶部１２２が管制局などに設置されている管制装置２０に搭載されることが考えられる。 As shown in FIG. 1, the environment information acquisition unit 101, the self-position estimation unit 105, the route generation unit 106, the movement control unit 107, the environment information storage unit 121, and the map storage unit 122 may be mounted on the autonomous mobile device 10. Conceivable. Similarly, it is conceivable that the environment information storage unit 121, the travel locus correction unit 102, the partial map generation unit 103, the map generation unit 104, and the map storage unit 122 are mounted on the control device 20 installed in a control station or the like. .
　なお、各部１０１～１０７，１２１，１２２は図１のように自律移動装置１０、管制装置２０に搭載されるとは限らず、例えば、各部１０１～１０７，１２１，１２２のすべてが自律移動装置１０に搭載されてもよいし、環境情報取得部１０１、走行軌跡補正部１０２、自己位置推定部１０５、経路生成部１０６、移動制御部１０７および環境情報記憶部１２１が自律移動装置１０に搭載されてもよい。 The units 101 to 107, 121, and 122 are not necessarily mounted on the autonomous mobile device 10 and the control device 20 as shown in FIG. 1. For example, all the units 101 to 107, 121, and 122 are all mounted on the autonomous mobile device 10. The environment information acquisition unit 101, the travel locus correction unit 102, the self-position estimation unit 105, the route generation unit 106, the movement control unit 107, and the environment information storage unit 121 are mounted on the autonomous mobile device 10. Also good.
　本実施形態では、図１に示すような構成であるものとする。そして、本実施形態では、自律移動装置１０で取得された環境情報が、管制装置２０へ送信される。そして、管制装置２０が環境情報に基づく走行軌跡の補正を行った後、部分地図を生成し、さらにその部分地図を接続して広域的な地図を生成する。そして、管制装置２０は、生成した地図を自律移動装置１０へ送る。自律移動装置１０は、送られた地図を基に自己位置を推定し、移動を行う。 In this embodiment, it is assumed that the configuration is as shown in FIG. In the present embodiment, the environmental information acquired by the autonomous mobile device 10 is transmitted to the control device 20. Then, after the control device 20 corrects the travel locus based on the environmental information, a partial map is generated, and the partial map is further connected to generate a wide-area map. Then, the control device 20 sends the generated map to the autonomous mobile device 10. The autonomous mobile device 10 estimates its own position based on the sent map and moves.
　図２は、本実施形態に係る自律移動装置および管制装置のハードウェア構成例を示す図である。
　図２（ａ）は、自律移動装置１０のハードウェア構成例を示す図である。
　自律移動装置１０は、ＣＰＵ（Central Processing Unit）２０１、ＲＯＭ（Read Only Memory）などのメモリ２０２、通信インタフェース２０３がバス２０４を介して互いに接続されている。
　メモリ２０２にはプログラムが格納されており、このプログラムをＣＰＵ２０１が実行することで、図１の自己位置推定部１０５、経路生成部１０６および移動制御部１０７などが具現化する。 FIG. 2 is a diagram illustrating a hardware configuration example of the autonomous mobile device and the control device according to the present embodiment.
FIG. 2A is a diagram illustrating a hardware configuration example of the autonomous mobile device 10.
In the autonomous mobile device 10, a CPU (Central Processing Unit) 201, a memory 202 such as a ROM (Read Only Memory), and a communication interface 203 are connected to each other via a bus 204.
A program is stored in the memory 202, and when the CPU 201 executes the program, the self-position estimation unit 105, the route generation unit 106, the movement control unit 107, and the like of FIG.
　図２（ｂ）は、管制装置２０のハードウェア構成例を示す図である。
　管制装置２０は、ＣＰＵ２１１、ＲＡＭ（Random Access Memory）２１２、ＲＯＭ２１３、通信インタフェース２１４、ＨＤ（Hard Disk）２１５がバス２１６を介して互いに接続されている。
　ＲＯＭ２１３や、ＨＤ２１５に格納されたプログラムが、ＲＡＭ２１２に展開され、このプログラムをＣＰＵ２１１が実行することによって、走行軌跡補正部１０２や、部分地図生成部１０３や、地図生成部１０４などが具現化する。 FIG. 2B is a diagram illustrating a hardware configuration example of the control device 20.
In the control device 20, a CPU 211, a RAM (Random Access Memory) 212, a ROM 213, a communication interface 214, and an HD (Hard Disk) 215 are connected to each other via a bus 216.
A program stored in the ROM 213 and the HD 215 is expanded in the RAM 212, and the CPU 211 executes the program, thereby realizing the travel locus correction unit 102, the partial map generation unit 103, the map generation unit 104, and the like.
［フローチャート］
　図３は、本実施形態に係る地図生成処理の手順を示すフローチャートである。なお、この処理は手動走行時に行われる処理である。
　まず、環境情報取得部１０１が走行路の部分領域に関する環境情報を取得する（Ｓ１０１）。
　そして、走行軌跡補正部１０２が環境情報の１つである移動座標に基づく走行軌跡を補正する走行軌跡補正処理を行う（Ｓ１０２）。ステップＳ１０２については、図５～図１６を参照して後記する。移動座標とは、デットレコニング、すなわちホイールオドメトリや、ジャイロオドメトリ、ビジュアルオドメトリなど（以下、単にオドメトリと呼ぶ）によって取得された座標データである。そして、走行軌跡は、各移動座標をつないだ軌跡であり、自律移動装置１０が移動した軌跡である。
　次に、部分地図生成部１０３が、補正した走行軌跡を基に、環境情報から得られる形状データを補完して部分地図を生成する（Ｓ１０３）。
　続いて、地図生成部１０４が、補正した走行軌跡を基に部分地図を接続して広域的な地図を生成し（Ｓ１０４）、地図記憶部１２２に格納する。ステップＳ１０４については、図１７、図１８を参照して後記する。 [flowchart]
FIG. 3 is a flowchart showing the procedure of map generation processing according to the present embodiment. This process is a process performed during manual travel.
First, the environment information acquisition unit 101 acquires environment information related to a partial area of the traveling road (S101).
Then, the travel locus correction unit 102 performs a travel locus correction process for correcting the travel locus based on the movement coordinates, which is one of the environment information (S102). Step S102 will be described later with reference to FIGS. The moving coordinates are coordinate data acquired by dead reckoning, that is, wheel odometry, gyro odometry, visual odometry (hereinafter simply referred to as odometry). And a driving | running | working locus | trajectory is a locus | trajectory which connected each movement coordinate, and is the locus | trajectory which the autonomous mobile device 10 moved.
Next, the partial map generation unit 103 generates a partial map by complementing the shape data obtained from the environment information based on the corrected travel locus (S103).
Subsequently, the map generation unit 104 connects the partial maps based on the corrected travel locus, generates a wide area map (S104), and stores the map in the map storage unit 122. Step S104 will be described later with reference to FIGS.
　図４は、本実施形態に係る自律移動処理の手順を示すフローチャートである。
　まず、環境情報取得部１０１が環境情報を取得する（Ｓ２０１）。
　次に、自己位置推定部１０５が、ステップＳ２０１で取得した環境情報と、地図記憶部１２２に格納している地図とを照合することにより自己位置を推定する（Ｓ２０２）。
　続いて、経路生成部１０６が、ステップＳ２０１で取得した環境情報から抽出した静止障害物や、移動障害物の情報や、自己位置の情報や、地図などを基に、自律移動装置１０が進む経路を生成する（Ｓ２０３）。
　そして、移動制御部１０７が、生成された経路に従って、自律移動装置１０を移動させる（Ｓ２０４）。 FIG. 4 is a flowchart showing the procedure of the autonomous movement process according to the present embodiment.
First, the environment information acquisition unit 101 acquires environment information (S201).
Next, the self-position estimation unit 105 estimates the self-position by collating the environment information acquired in step S201 with the map stored in the map storage unit 122 (S202).
Subsequently, the route traveled by the autonomous mobile device 10 based on the stationary obstacle, the moving obstacle information, the self-location information, the map, and the like extracted from the environment information acquired in step S201 by the route generation unit 106. Is generated (S203).
Then, the movement control unit 107 moves the autonomous mobile device 10 according to the generated route (S204).
　図５は、本実施形態に係る走行軌跡補正処理（図３のステップＳ１０２）の詳細な手順を示す図である。
　まず、走行軌跡補正部１０２は、環境情報の１つであるＧＰＳ測位情報などを用いて移動座標に基づいた走行軌跡の補正である第１の補正を行う（Ｓ３０１）。移動座標は、移動量情報と運動量情報(角速度や加速度、地磁気方位を指す)を積分することにより得られる。移動座標は、移動機構と路面とのスリップや、センサの計測誤差などの影響で誤差が蓄積する。つまり、移動座標では、移動距離が大きくなるほど、誤差が大きくなっていく。そして、前記したように各移動座標をつないだ軌跡が走行軌跡である。 FIG. 5 is a diagram showing a detailed procedure of the travel locus correction process (step S102 in FIG. 3) according to the present embodiment.
First, the travel locus correction unit 102 performs first correction, which is correction of the travel locus based on the movement coordinates, using GPS positioning information that is one of the environment information (S301). The movement coordinates are obtained by integrating movement amount information and momentum information (pointing to angular velocity, acceleration, and geomagnetic direction). The movement coordinates accumulate errors due to slippage between the movement mechanism and the road surface, sensor measurement error, and the like. That is, in the movement coordinates, the error increases as the movement distance increases. As described above, the trajectory connecting the movement coordinates is the travel trajectory.
　図６は、本実施形態に係る移動座標による走行軌跡の例を示す図である。
　図６の走行軌跡３００に示すように、本実施形態では、３次元オドメトリによる高さ方向の情報を有した３次元的な移動座標に基づく走行軌跡を想定している。なお、図６の走行軌跡３００は後記するようなオドメトリによる誤差などを考慮していない。走行軌跡３００は、実際には後記するような誤差を含んでいる。
　従って、走行軌跡補正部１０２は、図５のステップＳ３０１における走行軌跡を補正することで、移動座標の水平方向と高さ方向とを３次元的に補正する。
　なお、移動座標は、図６に示すような３次元的なものに限らず、２次元オドメトリによる２次元的なものでもよい。 FIG. 6 is a diagram illustrating an example of a travel locus based on movement coordinates according to the present embodiment.
As shown in a travel locus 300 in FIG. 6, in the present embodiment, a travel locus based on three-dimensional movement coordinates having information in the height direction based on three-dimensional odometry is assumed. Note that the travel locus 300 in FIG. 6 does not consider errors due to odometry, which will be described later. The travel locus 300 actually includes errors as described later.
Therefore, the traveling locus correction unit 102 corrects the traveling locus in step S301 in FIG. 5 to three-dimensionally correct the horizontal direction and the height direction of the movement coordinates.
The moving coordinates are not limited to the three-dimensional coordinates as shown in FIG. 6, and may be two-dimensional coordinates based on two-dimensional odometry.
　ステップＳ３０１の処理については、図８～図１１を参照して後記する。
　次に、図５の説明に戻り、走行軌跡補正部１０２は、自律移動装置１０が重複して走行する領域である重複走行領域の移動座標の対応付けを行う（Ｓ３０２）。ステップＳ３０２の処理については、図１２を参照して後記する。
　そして、走行軌跡補正部１０２は、ステップＳ３０２の結果を用いて、再度ステップＳ３０１と同様の処理を行い、走行軌跡を補正する第２の補正を行う（Ｓ３０３）。 The processing in step S301 will be described later with reference to FIGS.
Next, returning to the description of FIG. 5, the traveling locus correction unit 102 associates the movement coordinates of the overlapping traveling area, which is an area where the autonomous mobile device 10 travels in duplicate (S302). The process of step S302 will be described later with reference to FIG.
Then, the travel locus correction unit 102 performs the same process as in step S301 again using the result of step S302, and performs the second correction for correcting the travel locus (S303).
［走行領域］
　次に、図１を参照しつつ、図７～図１８を参照して、本実施形態に係る走行軌跡補正部１０２、部分地図生成部１０３、地図生成部１０４による走行軌跡補正と地図生成について詳細に説明する。なお、本実施形態では、図６に示すような３次元オドメトリによる３次元移動座標を用いることを想定しているが、説明を簡単にするため、図７～図１８では、２次元移動座標として説明する。
　図７は、本実施形態に係る自律移動装置の走行領域の例を示す図である。
　例えば、自律移動の対象とする走行路６０１の全体を含む走行領域６００は広域であるため、自律移動システム１は、部分領域６１０における環境情報を取得し、部分領域６１０を地図化した部分地図を生成する。図１８に示すように、この部分地図は異なる部分領域６１０，９０１，９０２で重複した領域が生じるよう複数生成される。そして、自律移動システム１は、生成される複数の部分地図を結合させることによって走行領域６００全体を地図化した走行地図を最終的に生成する。ここで、部分領域６１０は、自律移動装置１０が１回の走行で環境情報を取得する領域である。
　符号６１１～６１３，６２１，６２２については後記する。 [Running area]
Next, referring to FIGS. 7 to 18 with reference to FIG. 1, details of the travel locus correction and map generation by the travel locus correction unit 102, the partial map generation unit 103, and the map generation unit 104 according to the present embodiment will be described. Explained. In the present embodiment, it is assumed that three-dimensional movement coordinates by three-dimensional odometry as shown in FIG. 6 are used. However, in order to simplify the explanation, FIGS. 7 to 18 show two-dimensional movement coordinates. explain.
FIG. 7 is a diagram illustrating an example of a travel area of the autonomous mobile device according to the present embodiment.
For example, since the travel area 600 including the entire travel path 601 targeted for autonomous movement is a wide area, the autonomous mobile system 1 acquires environmental information in the partial area 610 and maps a partial map obtained by mapping the partial area 610. Generate. As shown in FIG. 18, a plurality of partial maps are generated so that overlapping regions are generated in different partial regions 610, 901, and 902. And the autonomous mobile system 1 finally produces | generates the driving | running | working map which mapped the whole driving | running | working area | region 600 by combining the some partial map produced | generated. Here, the partial area 610 is an area where the autonomous mobile device 10 acquires environment information in one run.
Reference numerals 611 to 613, 621, and 622 will be described later.
［第１の補正］
　図８は、本実施形態に係る走行軌跡の例を示す図である。
　図８の部分領域６１０において、自律移動装置１０が１回の走行（例えば手動走行）で経路７０１に沿って移動したとする。この走行で取得される真の移動座標は経路７０１上の移動座標となるはずである。つまり、走行軌跡は経路７０１の形状となるはずである。 [First correction]
FIG. 8 is a diagram illustrating an example of a travel locus according to the present embodiment.
In the partial area 610 of FIG. 8, it is assumed that the autonomous mobile device 10 has moved along the route 701 in a single travel (for example, manual travel). The true movement coordinates acquired by this traveling should be the movement coordinates on the route 701. That is, the travel locus should be the shape of the route 701.
　ここで、補正前の移動座標による走行軌跡を符号７１１で示す。
　環境情報取得部１０１で取得された移動座標は、前記したように、移動機構と路面とのスリップや、センサの計測誤差などの影響で誤差が蓄積していく。その結果、図８に示すように、環境情報取得部１０１で取得された移動座標に基づく走行軌跡７１１は、その形状に歪みが生じてしまう。つまり、環境情報取得部１０１で取得された移動座標に基づく走行軌跡７１１は、真の経路７０１との間に乖離が生じてしまう。 Here, a traveling locus based on the movement coordinates before correction is indicated by reference numeral 711.
As described above, errors are accumulated in the movement coordinates acquired by the environment information acquisition unit 101 due to the effects of slippage between the movement mechanism and the road surface, measurement errors of the sensor, and the like. As a result, as shown in FIG. 8, the travel locus 711 based on the movement coordinates acquired by the environment information acquisition unit 101 is distorted in shape. That is, the travel locus 711 based on the movement coordinates acquired by the environment information acquisition unit 101 is deviated from the true route 701.
　そこで、本実施形態において、走行軌跡補正部１０２は、まず初めにＧＰＳ測位情報（以下、より具体的にＧＰＳ座標とする）に基づく拘束によって、補正前の移動座標に基づく走行軌跡７１１を補正する。ＧＰＳ座標に基づく拘束は、移動座標の各地点の座標が、任意の絶対座標系（例えば、緯度、経度、標高などの世界座標系）において、どこに相当するのかを示すものである。
　そして、走行軌跡補正部１０２は、補正前の移動座標に基づく走行軌跡７１１を変形させ、ＧＰＳ座標との乖離が最も小さくなる走行軌跡の形状を取得することで、移動座標を補正する。 Therefore, in the present embodiment, the traveling locus correction unit 102 first corrects the traveling locus 711 based on the uncorrected movement coordinates by the constraint based on the GPS positioning information (hereinafter, more specifically referred to as GPS coordinates). . The constraint based on the GPS coordinate indicates where the coordinate of each point of the movement coordinate corresponds to in an arbitrary absolute coordinate system (for example, a world coordinate system such as latitude, longitude, and altitude).
Then, the travel locus correction unit 102 corrects the travel coordinates by deforming the travel locus 711 based on the travel coordinates before correction and acquiring the shape of the travel locus that minimizes the deviation from the GPS coordinates.
　しかし、すべての拘束を完全に満たす変形は存在しないため、走行軌跡補正部１０２は、各拘束の物理的な意味に基づいて生じうる誤差を基に、例えば誤差を密度関数で定義する。そして、走行軌跡補正部１０２は、この密度関数に基づいた値が最大化または最小化するように移動座標の位置関係を変形することで、すべての拘束を最大限に満たしたと見なすようにしてもよい。つまり、走行軌跡補正部１０２は、最適化処理によって、すべての拘束を最大限に満たしたとみなす。走行軌跡の位置関係の変形手順には、例えばＧｒａｐｈ－ＳＬＡＭ（Simultaneous Location And Mapping）法をはじめとする連立方程式の最適化問題として解く手法が用いられてもよいし、走行軌跡に対してランダムな微小補正をトライ＆エラーで重畳してゆくことで拘束を満たす手順による手法が用いられてもよい。
　ここで、図９～図１１を参照して、補正前の走行軌跡７１１を変形させて、補正後の走行軌跡を得る具体的な手順を詳細に説明する。 However, since there is no deformation that completely satisfies all constraints, the travel locus correction unit 102 defines, for example, the error as a density function based on errors that can occur based on the physical meaning of each constraint. Then, the traveling locus correction unit 102 may consider that all constraints are satisfied to the maximum by changing the positional relationship of the movement coordinates so that the value based on the density function is maximized or minimized. Good. That is, the travel locus correction unit 102 considers that all constraints are satisfied to the maximum by the optimization process. For the transformation procedure of the positional relationship of the traveling locus, for example, a method of solving as an optimization problem of simultaneous equations such as a Graph-SLAM (Simultaneous Location And Mapping) method may be used, or a random relationship with respect to the traveling locus may be used. A technique based on a procedure that satisfies constraints by superimposing minute corrections by trial and error may be used.
Here, with reference to FIG. 9 to FIG. 11, a specific procedure for deforming the travel locus 711 before correction to obtain the corrected travel locus will be described in detail.
　図９は、本実施形態に係る第１の補正の詳細な手順を示す図である。
　図９に示すように、走行軌跡の補正を行う前の移動座標７２１（適宜、補正前の移動座標７２１と称する）による走行軌跡７１１が得られているものとする。
　これに対し、ＧＰＳ座標が符号７３１として取得されているものとする。なお、図９において「・」印がオドメトリによって測定された移動座標を示し、「×」印がＧＰＳによる座標（ＧＰＳ座標）を示している。
　ここで、補正前の移動座標７２１と、ＧＰＳ座標７３１との各々は対応付けられている（例えば、補正前の移動座標７２１ａとＧＰＳ座標７３１ａ、補正前の移動座標７２１ｂとＧＰＳ座標７３１ｂのように）。つまり、環境情報取得部１０１は、自律移動装置の走行軌跡を生成するための情報（移動座標）および前記走行軌跡の所定の位置に対応づけられた絶対座標（ＧＰＳ座標）を取得する。 FIG. 9 is a diagram showing a detailed procedure of the first correction according to the present embodiment.
As shown in FIG. 9, it is assumed that a travel locus 711 based on a movement coordinate 721 before correction of the travel locus (referred to as a movement coordinate 721 before correction as appropriate) is obtained.
On the other hand, it is assumed that GPS coordinates are acquired as reference numeral 731. In FIG. 9, “·” marks indicate movement coordinates measured by odometry, and “x” marks indicate GPS coordinates (GPS coordinates).
Here, the movement coordinates 721 before correction and the GPS coordinates 731 are associated with each other (for example, movement coordinates 721a and GPS coordinates 731a before correction, movement coordinates 721b and GPS coordinates 731b before correction, and the like). ). That is, the environment information acquisition unit 101 acquires information (movement coordinates) for generating a traveling locus of the autonomous mobile device and absolute coordinates (GPS coordinates) associated with a predetermined position of the traveling locus.
　そして、走行軌跡補正部１０２は、補正前の移動座標７２１と、ＧＰＳ座標７３１とを基に、走行軌跡７１１を変形することによって、破線で示す補正後の走行軌跡を得る。走行軌跡補正部１０２は、例えば、補正前の移動座標７２１とＧＰＳ座標７３１とから定義される誤差評価値を最小にする補正後の走行軌跡７４１を生成する。そして、このような走行軌跡７１１の補正に伴って走行軌跡７１１上の移動座標７２１も補正されることとなる。 The traveling locus correction unit 102 obtains a corrected traveling locus indicated by a broken line by deforming the traveling locus 711 based on the movement coordinates 721 before correction and the GPS coordinates 731. For example, the travel locus correction unit 102 generates a corrected travel locus 741 that minimizes an error evaluation value defined from the movement coordinates 721 and the GPS coordinates 731 before correction. As the travel locus 711 is corrected, the movement coordinates 721 on the travel locus 711 are also corrected.
　図１０は、本実施形態に係る第１の補正に用いられる誤差評価値（第１の誤差評価値）の一例を示す図である。ここで、図１０は、図９に示す走行軌跡の一部を拡大したものであり、図９と同一の構成については、同一の符号を付している。なお、補正前の移動座標７２１（以下、適宜、移動座標７２１と称する）にＧＰＳ座標７３１が対応付けられていない地点は、ＧＰＳ座標が取得できなかった地点である。 FIG. 10 is a diagram illustrating an example of an error evaluation value (first error evaluation value) used for the first correction according to the present embodiment. Here, FIG. 10 is an enlarged view of a part of the travel locus shown in FIG. 9, and the same components as those in FIG. 9 are denoted by the same reference numerals. A point where the GPS coordinate 731 is not associated with the uncorrected movement coordinate 721 (hereinafter referred to as the movement coordinate 721 as appropriate) is a point where the GPS coordinate could not be acquired.
　まず、走行軌跡補正部１０２は取得されたＧＰＳ座標７３１からの距離に対する密度関数で定義されるＧＰＳ座標の存在確率を求め、その標準偏差（σ）に２を乗算した値（２σ）が示す領域（２σ領域７３２）を各々のＧＰＳ座標７３１について算出する。
　そして、走行軌跡補正部１０２は２σ領域７３２のうちで移動座標７２１に最も近い点から移動座標７２１に線分７３３をひく。走行軌跡補正部１０２は、この線分７３３を、それぞれの移動座標７２１およびＧＰＳ座標７３１の対について算出する。このとき、ＧＰＳ座標７３１が取得できていない地点は、線分７３３の長さを「０」とする。 First, the travel locus correction unit 102 obtains the existence probability of a GPS coordinate defined by a density function with respect to the distance from the acquired GPS coordinate 731, and an area indicated by a value (2σ) obtained by multiplying the standard deviation (σ) by 2 (2σ region 732) is calculated for each GPS coordinate 731.
Then, the travel locus correction unit 102 draws a line segment 733 from the point closest to the movement coordinate 721 in the 2σ region 732 to the movement coordinate 721. The travel locus correction unit 102 calculates the line segment 733 for each pair of movement coordinates 721 and GPS coordinates 731. At this time, the length of the line segment 733 is set to “0” at a point where the GPS coordinate 731 cannot be acquired.
　そして、走行軌跡補正部１０２は、各々の線分７３３の２乗和を誤差評価値とし、この２乗和が最も小さくなるよう補正前の走行軌跡７１１（適宜、走行軌跡７１１と称する）を変形することによって、補正後の走行軌跡７４１（図９）を得る。このとき、曲がり角（例えば、走行軌跡７１１のなす各が一定の大きさ以上の箇所）以外では、走行軌跡７１１ができるだけ直線性を保つように走行軌跡を変形するとよい。このとき、走行軌跡補正部１０２は、走行軌跡７１１を少し変形させて誤差評価値としての線分７３３の合計値を求め、次に、走行軌跡７１１をさらに少し変形させて線分７３３の合計値を求めることを繰り返して、線分７３３の合計値が最も小さくなる補正後の走行軌跡７４１を求める。
　このような手法を用いる理由は、ＧＰＳ座標７３１も誤差を含んでいるためである。 Then, the traveling locus correction unit 102 sets the square sum of each line segment 733 as an error evaluation value, and deforms the traveling locus 711 before correction (referred to as traveling locus 711 as appropriate) so that the square sum becomes the smallest. By doing so, a corrected travel locus 741 (FIG. 9) is obtained. At this time, it is preferable that the travel locus is deformed so that the travel locus 711 is as linear as possible except for a corner (for example, a portion where each of the travel locus 711 has a certain size or more). At this time, the traveling locus correction unit 102 slightly deforms the traveling locus 711 to obtain a total value of the line segment 733 as an error evaluation value, and then deforms the traveling locus 711 further slightly to obtain the total value of the line segment 733. Is repeated to obtain a corrected travel locus 741 in which the total value of the line segment 733 is the smallest.
The reason for using such a method is that the GPS coordinate 731 also includes an error.
　なお、一般に用いられる高精度なＧＰＳ測位であるＤ－ＧＰＳ（Differential-GPS）とＲＴＫ－ＧＰＳ（Real Time Kinematic-GPS）のうち、ＲＴＫ－ＧＰＳは精度が非常に高いが、測位には衛星が５個以上必要、かつ、一度測位に失敗すると再測位をするまでの間、自律移動装置１０を静止させておく必要があるため簡便に用いることができない。これに対し、Ｄ－ＧＰＳは自律移動装置１０が移動中でも再測位が可能であり、必要な衛星数が少ないため簡便に用いることができる。 Of D-GPS (Differential-GPS) and RTK-GPS (Real Time Kinematic-GPS), which are commonly used high-accuracy GPS positioning, RTK-GPS has a very high accuracy, but positioning requires a satellite. Since it is necessary to keep the autonomous mobile device 10 stationary until five or more pieces are necessary and the positioning is once failed, it cannot be used easily. On the other hand, D-GPS can be re-positioned while the autonomous mobile device 10 is moving, and can be easily used because the number of necessary satellites is small.
　ただし、ＧＰＳ座標は一般に緯度・経度からなる位置と比べて高度(標高)の計測精度が低い。このため、移動座標の高さ情報を補正するのに十分ではないことが多い。そこで、本実施形態では、例えば航空機などで事前に測量した部分領域６１０（図７）の高さ情報をＧＰＳ座標と共に用いることで移動座標の高さを補正するようにしてもよい。その手法を図１１を参照して説明する。 However, GPS coordinates are generally less accurate in measuring altitude (elevation) than positions consisting of latitude and longitude. For this reason, it is often not sufficient to correct the height information of the moving coordinates. Therefore, in the present embodiment, the height of the moving coordinate may be corrected by using the height information of the partial area 610 (FIG. 7) measured in advance with an aircraft or the like together with the GPS coordinates. The method will be described with reference to FIG.
　ここでは、ＧＰＳ座標７３１とともに、例えば航空機から取得した走行路の高さ情報を用いることで、走行軌跡の補正を行うことを説明する。
　このような高さ情報は、航空機によるレーダなどを用いた高さ情報である。
　図１１は、ＧＰＳ座標７３１に加えて、航空機から取得した走行路の高さ情報（以下、航空機情報７５１と称する）を考慮した補正に用いる誤差評価値の一例を示す図である。なお、図１１で用いる構成について、図１０と同様の構成については、同一の符号を付して説明を省略する。
　なお、図１１において、航空機情報７５１がない場所は屋根や、木などによって航空機情報７５１を取得することができなかった場所である。 Here, correction of the travel locus will be described by using, for example, the height information of the travel route acquired from the aircraft together with the GPS coordinates 731.
Such height information is height information using an aircraft radar or the like.
FIG. 11 is a diagram illustrating an example of an error evaluation value used for correction in consideration of the height information of the traveling path acquired from the aircraft (hereinafter referred to as aircraft information 751) in addition to the GPS coordinates 731. In addition, about the structure used in FIG. 11, about the structure similar to FIG. 10, the same code | symbol is attached | subjected and description is abbreviate | omitted.
In FIG. 11, a place without the aircraft information 751 is a place where the aircraft information 751 could not be acquired due to a roof or a tree.
　ここで、走行軌跡補正部１０２は、航空機情報７５１について、図１０におけるＧＰＳ座標７３１と同様に２σ領域７５２を算出し、その２σ領域７５２のうちで補正前の移動座標７２１（適宜、移動座標７２１と称する）に最も近い点から移動座標７２１に線分７５３をひく。走行軌跡補正部１０２は、この線分７５３を、それぞれの移動座標７２１および航空機情報７５１の対について算出する。このとき、航空機座標７５１が取得できていない地点では、線分７５３の長さは「０」となる。 Here, the travel locus correction unit 102 calculates a 2σ region 752 for the aircraft information 751 in the same manner as the GPS coordinate 731 in FIG. 10, and the uncorrected movement coordinate 721 (appropriately, the movement coordinate 721 in the 2σ region 752). A line segment 753 is drawn on the movement coordinate 721 from the point closest to the above. The travel locus correction unit 102 calculates the line segment 753 for each pair of movement coordinates 721 and aircraft information 751. At this time, the length of the line segment 753 is “0” at a point where the aircraft coordinates 751 cannot be acquired.
　そして、走行軌跡補正部１０２は、すべての線分７３３とすべての線分７５３の合計値を誤差評価値（第１の誤差評価値）とし、この合計値が最も小さくなるよう補正前の走行軌跡７１１（適宜、走行軌跡７１１と称する）を変形する。これにより、走行軌跡補正部１０２は、補正後の走行軌跡７４１（図９）を得る。このとき、図１０と同様、曲がり角（例えば、走行軌跡７１１のなす各が一定の大きさ以上の箇所）以外では、走行軌跡７１１ができるだけ直線性を保つように走行軌跡を変形するとよい。
　ここで、走行軌跡補正部１０２は、ＧＰＳ座標７３１と、航空機情報７５１の精度に応じて重み付けを加えてもよい。
　例えば、航空機情報７５１の方がＧＰＳ座標７３１より精度が高ければ、走行軌跡補正部１０２は線分７５３の重み付けを大きくする。 Then, the traveling locus correction unit 102 sets the total value of all the line segments 733 and all the line segments 753 as an error evaluation value (first error evaluation value), and the traveling locus before correction so that the total value becomes the smallest. 711 (referred to as travel locus 711 as appropriate) is modified. Thereby, the traveling locus correction unit 102 obtains a corrected traveling locus 741 (FIG. 9). At this time, similarly to FIG. 10, the travel locus may be deformed so that the travel locus 711 is as linear as possible except at a corner (for example, each portion formed by the travel locus 711 has a certain size or more).
Here, the traveling locus correction unit 102 may add weight according to the accuracy of the GPS coordinates 731 and the aircraft information 751.
For example, if the aircraft information 751 is more accurate than the GPS coordinate 731, the traveling locus correction unit 102 increases the weight of the line segment 753.
　なお、図１０、図１１で示した走行軌跡の補正は一例であり、他の手法が使用されてもよい。
　例えば、遺伝的アルゴリズムや、ニューラルネットワークなどを用いることによって、補正後の走行軌跡７４１が求められてもよい。
　以上が、図５におけるステップＳ３０１における走行軌跡補正の処理である。 Note that the correction of the travel locus shown in FIGS. 10 and 11 is an example, and other methods may be used.
For example, the corrected travel locus 741 may be obtained by using a genetic algorithm, a neural network, or the like.
The above is the travel locus correction processing in step S301 in FIG.
　従来のように移動座標を逐次的に補正すると、トンネルなどＧＰＳ座標が取得されない場所では、移動座標の補正がなされないので、その間、誤差が蓄積されてしまう。そして、ＧＰＳ座標が取得可能なところで、それまでの誤差が蓄積されている状態から、突然、補正がなされるので、走行軌跡が不連続的になってしまう。
　これに対し、本実施形態のような移動座標の補正を行うと、他の場所の影響を考慮して、全体的に走行軌跡を補正するため、ＧＰＳ座標が取得されない場所からＧＰＳ座標が取得される場所へと移ったところでも、連続的な走行軌跡を得ることができる。 If the movement coordinates are sequentially corrected as in the prior art, the movement coordinates are not corrected at locations where GPS coordinates are not acquired, such as tunnels, and errors are accumulated during that time. And since GPS coordinates can be acquired, the correction is made suddenly from the state in which the errors up to that point are accumulated, so that the travel locus becomes discontinuous.
On the other hand, when the movement coordinates are corrected as in the present embodiment, the GPS coordinates are acquired from a place where the GPS coordinates are not acquired in order to correct the travel locus in consideration of the influence of other places. A continuous running track can be obtained even when the vehicle moves to a place.
　このようにして、走行軌跡補正部１０２は、絶対座標に基づく第１の誤差評価値を算出し、相対座標により得られる自律移動装置の走行軌跡を、第１の誤差評価値に基づいて変形させることで、走行軌跡を補正する第１の補正を行う。 In this way, the travel locus correction unit 102 calculates the first error evaluation value based on the absolute coordinates, and deforms the travel locus of the autonomous mobile device obtained from the relative coordinates based on the first error evaluation value. Thus, the first correction for correcting the travel locus is performed.
［重複走行領域における移動座標の対応付け］
　このように、まず図５のステップＳ３０１の処理によって、おおまかな走行軌跡の補正が行われる。
　図１２は、重複走行領域の移動座標の対応付け（図５のステップＳ３０２）の詳細な手順を説明する図である。
　ステップＳ３０１による走行軌跡補正の結果、走行軌跡補正部１０２は、ステップＳ３０１で生成された補正後の走行軌跡７４１のうちで、自律移動装置１０が図１２（ａ）に示すように近傍を通っていると推測できる領域（重複走行領域７７１）を得ることができる。つまり、走行軌跡補正部１０２は、補正後の走行軌跡が近接している領域を検出することによって、重複走行領域７７１の検出を行う。つまり、走行軌跡補正部１０２は、第１の補正による補正が行われた走行軌跡を構成する軌跡同士が沿うように近接している領域を検出することによって、自律移動装置が重複して走行した領域である重複走行領域の検出を行う [Association of moving coordinates in overlapping travel areas]
In this way, first, the rough travel locus is corrected by the processing in step S301 in FIG.
FIG. 12 is a diagram for explaining the detailed procedure of associating the movement coordinates of the overlapping travel area (step S302 in FIG. 5).
As a result of the travel trajectory correction in step S301, the travel trajectory correction unit 102 causes the autonomous mobile device 10 to pass through the vicinity of the corrected travel trajectory 741 generated in step S301 as shown in FIG. It is possible to obtain a region (overlapping traveling region 771) that can be estimated as being. That is, the traveling locus correction unit 102 detects the overlapping traveling region 771 by detecting a region where the corrected traveling locus is close. In other words, the travel locus correction unit 102 has detected that the regions that are close to each other so that the tracks constituting the travel locus that has been corrected by the first correction are located, so that the autonomous mobile devices have traveled in duplicate. Detect overlapping running areas that are areas
　ここで、走行軌跡補正部１０２は、例えば、移動座標７６１，７６２それぞれの地点で得られる周囲の形状データを比較することで、移動座標７６１，７６２の相対的な位置関係を算出する。ここで、形状データは、例えば、レーザ距離センサで取得される建物などの形状データである。なお、形状データは、環境情報に含まれる情報である。 Here, the traveling locus correction unit 102 calculates the relative positional relationship between the movement coordinates 761 and 762, for example, by comparing the surrounding shape data obtained at the respective points of the movement coordinates 761 and 762. Here, shape data is shape data, such as a building acquired with a laser distance sensor, for example. The shape data is information included in the environment information.
　図１２（ｂ）、図１２（ｃ）を参照して、重複走行領域の移動座標の対応付けの具体的な手法を説明する。
　例えば、走行軌跡補正部１０２は、図１２（ｂ）に示すような、移動座標７６１において自律移動装置１０が取得した形状データ（図１２（ｂ）の太線部分７６５）と、図１２（ｃ）に示すような、移動座標７６２において自律移動装置１０が取得した形状データ（図１２（ｃ）の太線部分７６６）と、を比較する。この比較によって、重複走行領域７７１における移動座標間の対応付けが行われ、重複走行領域７７１における移動座標間の拘束が確定される。つまり、走行軌跡補正部１０２は、重複走行領域における走行軌跡上の相対座標を複数選択し、選択された各相対座標に対応付けられている環境情報同士を互いに比較することによって、選択された各相対座標間における第１の相対的な位置関係を算出する。
　具体的には、ステップＳ３０２の処理によって、移動座標７６１，７６２の相対的な位置関係が図１２（ｄ）に示すような位置関係（図１２（ｄ）の破線矢印）であることが算出される。 With reference to FIG. 12B and FIG. 12C, a specific method of associating the movement coordinates of the overlapping travel area will be described.
For example, as shown in FIG. 12B, the travel locus correction unit 102 obtains the shape data (the thick line portion 765 in FIG. 12B) acquired by the autonomous mobile device 10 at the movement coordinates 761, and FIG. The shape data (the thick line portion 766 in FIG. 12C) acquired by the autonomous mobile device 10 at the movement coordinates 762 as shown in FIG. By this comparison, the movement coordinates in the overlapping traveling area 771 are associated with each other, and the restriction between the moving coordinates in the overlapping traveling area 771 is determined. In other words, the traveling locus correction unit 102 selects a plurality of relative coordinates on the traveling locus in the overlapping traveling region, and compares the environment information associated with each selected relative coordinate with each other to select each of the selected relative coordinates. A first relative positional relationship between relative coordinates is calculated.
Specifically, it is calculated by the processing in step S302 that the relative positional relationship between the movement coordinates 761 and 762 is a positional relationship as shown in FIG. 12D (broken arrows in FIG. 12D). The
　また、図５のステップＳ３０２において、重複走行領域７７１において対応付けられた２つの移動座標間における拘束（相対的な位置関係）は、一方の移動座標（例えば、符号７６１）と他方の移動座標（例えば、符号７６２）とにおける自律移動装置１０の位置・姿勢が、どのような値を満たすべきかという意味を持つ。 Further, in step S302 of FIG. 5, the constraint (relative positional relationship) between the two movement coordinates associated with each other in the overlapping travel area 771 is that one movement coordinate (for example, reference numeral 761) and the other movement coordinate ( For example, it has a meaning as to what value the position / posture of the autonomous mobile device 10 at 762) should satisfy.
　ここで、走行軌跡補正部１０２は、例えば、環境情報取得部１０１から得られる環境情報における形状や模様を、点群処理におけるＩＣＰ(Iterative Closest Point)法や画像処理におけるテンプレートマッチング法により対応付けることで、移動座標間の対応付けを行う。 Here, for example, the travel locus correction unit 102 associates the shape or pattern in the environment information obtained from the environment information acquisition unit 101 by an ICP (Iterative Closest Point) method in point cloud processing or a template matching method in image processing. Correspondence between moving coordinates is performed.
　このようにすることで、走行軌跡補正部１０２は、相対座標系の拘束も満たすように移動座標を補正することができる。また、仮にＧＰＳ座標に基づく拘束にばらつき誤差が生じた際にも、同一地点の情報がずれた箇所に重複して配置されることのない移動座標を得ることができる。つまり、図５のステップＳ３０１で行われる走行軌跡補正の結果においても誤差が完全に消えるわけではない。しかしながら、このように補正された移動座標に誤差が含まれている場合でも、重複して走行した地点を別の地点として認識することがなくなる。 In this way, the traveling locus correction unit 102 can correct the movement coordinates so as to satisfy the constraints of the relative coordinate system. In addition, even when a variation error occurs in the constraint based on the GPS coordinates, it is possible to obtain moving coordinates that are not repeatedly arranged at a location where the information on the same point is shifted. That is, the error does not disappear completely even in the result of the travel locus correction performed in step S301 in FIG. However, even when the movement coordinates corrected in this way include an error, the overlapping traveling point is not recognized as another point.
　図５のステップＳ３０１において、走行軌跡補正部１０２は、初めにＧＰＳ座標や、航空機情報を用いて、走行軌跡７１１（移動座標）から誤差の蓄積を排除することにより、走行軌跡の大まかな補正を、まず行う。このようにすることで、移動座標が持つ誤差の最大値をＧＰＳ座標の誤差程度に制限することが可能となる。さらに、一回の走行における重複走行領域において、対応付けられた各地点の誤差もＧＰＳ座標の誤差程度に抑えることができる。従って、部分地図を結合して広域的な地図を生成する際の確実性を向上させることができる。 In step S301 in FIG. 5, the travel locus correction unit 102 first performs rough correction of the travel locus by eliminating the accumulation of errors from the travel locus 711 (moving coordinates) using GPS coordinates and aircraft information. First, do it. By doing in this way, it becomes possible to restrict the maximum value of the error that the moving coordinate has to the extent of the GPS coordinate error. Furthermore, in the overlapping traveling area in one traveling, the error at each associated point can be suppressed to the extent of the GPS coordinate error. Therefore, the certainty at the time of producing | generating a wide area map by combining a partial map can be improved.
[第２の補正]
　そして、走行軌跡補正部１０２は、重複走行領域７７１の対応付けを行った移動座標を用いて、再度ＧＰＳ座標を用いた第２の補正を行う（図５のステップＳ３０３）。つまり、走行軌跡補正部１０２は、算出した第１の相対的な位置関係と、絶対座標と、の両者に基づく第２の誤差評価値を算出し、走行軌跡を、当該第２の誤差評価値に基づいて変形させることで、走行軌跡を補正する第２の補正を行う。
　このとき、走行軌跡７１１は、ステップＳ３０１と同様の手順で再度補正される。 [Second correction]
Then, the traveling locus correction unit 102 performs the second correction using the GPS coordinates again using the movement coordinates in which the overlapping traveling regions 771 are associated (step S303 in FIG. 5). That is, the travel locus correction unit 102 calculates a second error evaluation value based on both the calculated first relative positional relationship and the absolute coordinates, and uses the travel locus as the second error evaluation value. By performing the deformation based on the second correction, the second correction for correcting the travel locus is performed.
At this time, the travel locus 711 is corrected again by the same procedure as in step S301.
　図１３は、本実施形態に係る第２の補正に用いられる誤差評価値（第２の誤差評価値）の一例を示す図である。図１３において、図１０と同様の構成要素については、同一の符号を付して説明を省略する。
　走行軌跡補正部１０２は、ステップＳ３０２で移動座標７２１に対応付けられた移動座標７８１から、移動座標７２１に線分７８３をひく。このとき、線分７８３の長さ・方向が、ステップＳ３０２で算出された移動座標７２１，７８１間の相対的な位置関係となるようにひかれる。走行軌跡補正部１０２は、この線分７８３を、それぞれの移動座標７２１について算出する。 FIG. 13 is a diagram illustrating an example of an error evaluation value (second error evaluation value) used for the second correction according to the present embodiment. In FIG. 13, the same components as those in FIG. 10 are denoted by the same reference numerals and description thereof is omitted.
The travel locus correction unit 102 draws a line segment 783 from the movement coordinate 781 associated with the movement coordinate 721 in step S302 to the movement coordinate 721. At this time, the length and direction of the line segment 783 are drawn so as to be a relative positional relationship between the movement coordinates 721 and 781 calculated in step S302. The travel locus correction unit 102 calculates the line segment 783 for each movement coordinate 721.
　そして、走行軌跡補正部１０２は、すべての線分７３３とすべての線分７８３の合計値を誤差評価値とし、この合計値が最も小さくなるよう第１の補正後の走行軌跡７４１を変形することで、第２の補正を行う。このとき、線分７８３の重み付けを大きくすることが望ましい（図１３では太線で示している）。また、図１０と同様、曲がり角（例えば、走行軌跡７１１のなす各が一定の大きさ以上の箇所）以外では、走行軌跡７１１ができるだけ直線性を保つように走行軌跡を変形するとよい。 Then, the travel locus correction unit 102 uses the total value of all the line segments 733 and all the line segments 783 as an error evaluation value, and deforms the travel locus 741 after the first correction so that the total value is minimized. Then, the second correction is performed. At this time, it is desirable to increase the weight of the line segment 783 (indicated by a thick line in FIG. 13). Similarly to FIG. 10, the travel locus may be deformed so that the travel locus 711 is as linear as possible except at a corner (for example, each of the travel locus 711 is a certain size or more).
　なお、本実施形態では、形状データに基づいて算出された移動座標間の相対的な位置関係には、ほとんど誤差が含まれないものとして、移動座標７８１と、移動座標７２１との間に線分を生成しているが、図１０、図１１と同様に、移動座標７８１の２σ領域を算出してもよい。そして、走行軌跡補正部１０２は、その２σ領域と移動座標７８１との間に線分７８３をひいて、すべての線分７３３とすべての線分７８３の合計値を誤差評価値としてもよい。 In the present embodiment, it is assumed that the relative positional relationship between the movement coordinates calculated based on the shape data includes almost no error, and a line segment between the movement coordinates 781 and the movement coordinates 721 is assumed. However, the 2σ region of the moving coordinate 781 may be calculated as in FIGS. 10 and 11. Then, the traveling locus correction unit 102 may draw a line segment 783 between the 2σ region and the movement coordinate 781 and use the total value of all the line segments 733 and all the line segments 783 as an error evaluation value.
　また、図１３において、相関的な位置関係で対応付けられている移動座標は、各異動座標毎に一対となっているが、複数の移動座標間の対応付けが使用されてもよい。例えば、移動座標７２１ａと、移動座標７８１ａ，７８１ｂ、７８１ｃ、７８１ｄそれぞれとの相対的な位置関係に基づいた長さの線分、移動座標７２１ｂと、移動座標７８１ａ，７８１ｂ、７８１ｃ、７８１ｄそれぞれとの相対的な位置関係に基づいた長さの線分などを誤差評価値に使用してもよい。
　なお、ステップＳ３０３において、走行軌跡補正部１０２は、ステップＳ３０１の結果を用いるのではなく、最初から（つまり、補正後の走行軌跡を用いるのではなく、補正前の走行軌跡を用いる）処理を行うことが望ましい。 In FIG. 13, the movement coordinates associated with each other in a relative positional relationship are paired for each transfer coordinate, but association between a plurality of movement coordinates may be used. For example, a line segment having a length based on the relative positional relationship between the moving coordinates 721a and the moving coordinates 781a, 781b, 781c, and 781d, the moving coordinates 721b, and the moving coordinates 781a, 781b, 781c, and 781d, respectively. A line segment having a length based on the relative positional relationship may be used as the error evaluation value.
In step S303, the travel locus correction unit 102 does not use the result of step S301, but performs processing from the beginning (that is, uses the travel locus before correction instead of using the corrected travel locus). It is desirable.
　また、本実施形態で用いられる走行軌跡の補正に用いられる環境情報を、自律移動装置１０で取得した移動座標、ＧＰＳ座標および航空機による環境情報の３種類としているが、この３種類に限らず、４種類以上でもよい。また、絶対座標であれば、ＧＰＳ座標や、航空機による環境情報以外の情報が用いられてもよい。なお、重複走行領域がなるべく多くなるよう自律移動装置１０の走行経路を予め設定しておくことが望ましい。 Moreover, although the environment information used for correction | amendment of the driving | running | working locus | trajectory used by this embodiment is made into three types of the movement information acquired by the autonomous mobile device 10, GPS coordinates, and the environment information by an aircraft, it is not restricted to these three types, There may be four or more types. Moreover, as long as it is an absolute coordinate, information other than the GPS coordinate or the environment information by the aircraft may be used. In addition, it is desirable to set the traveling route of the autonomous mobile device 10 in advance so that the overlapping traveling area is as large as possible.
　図１４は、本実施形態に係る第１の補正、第２の補正に伴う走行軌跡の変化を示す図である。
　図１４（ａ）における走行軌跡７１１は、第１の補正前の走行軌跡であり、オドメトリから取得された移動座標そのものである。
　この走行軌跡７１１に対し、図９～図１１で示した第１の補正が行われることで、図１４（ｂ）に示す７４１が得られる。走行軌跡７４１は、走行軌跡７１１に比べ、真の走行軌跡（図８の経路７０１の形状）に近いものとなっているが、ＧＰＳ座標などの絶対座標における誤差の影響が残っている。言い換えれば、走行軌跡７４１における誤差は、ＧＰＳ座標などの絶対座標における誤差程度に抑えられている。
　そして、走行軌跡補正部１０２は、第１の補正後の走行軌跡７４１に対し重複走行領域を検出し、図１２、図１３で示した第２の補正が行われることで、図１４（ｃ）に示す走行軌跡７９１が得られる。ＧＰＳ座標などよりも誤差が小さい重複走行領域における移動座標の相対的な位置関係を用いることで、走行軌跡７９１は、走行軌跡７１１，７４１よりも真の走行軌跡（図８の経路７０１の形状）に近いものとなっている。 FIG. 14 is a diagram illustrating a change in the travel locus associated with the first correction and the second correction according to the present embodiment.
A travel trajectory 711 in FIG. 14A is a travel trajectory before the first correction, and is a movement coordinate itself acquired from odometry.
741 shown in FIG. 14B is obtained by performing the first correction shown in FIGS. 9 to 11 on the travel locus 711. The travel locus 741 is closer to the true travel locus (the shape of the route 701 in FIG. 8) than the travel locus 711, but the influence of errors in absolute coordinates such as GPS coordinates remains. In other words, the error in the travel locus 741 is suppressed to the extent of an error in absolute coordinates such as GPS coordinates.
Then, the travel locus correction unit 102 detects an overlapping travel region for the travel locus 741 after the first correction, and the second correction shown in FIGS. 12 and 13 is performed, so that FIG. A travel locus 791 shown in FIG. By using the relative positional relationship of the movement coordinates in the overlapping traveling area where the error is smaller than that of the GPS coordinates or the like, the traveling locus 791 is a true traveling locus (the shape of the route 701 in FIG. 8) than the traveling locus 711 or 741 It is close to.
［上空から取得した環境情報の使用の可・不可の判定］
　図７に示されているように、自律移動の対象である走行路６０１は、周囲に木や建物などの構造物６１１が存在する領域６１２と、周囲に構造物が存在しない領域６１３が含まれる。なお、図７の領域６１２と領域６１３それぞれからの吹き出し６２１，６２２は各領域６１２，６１３の鳥瞰図を示す。ここで、航空機で取得される環境情報を上空環境情報と称することとする。前記した航空機情報（航空機から取得した走行路の高さ情報）は、上空環境情報に含まれる情報である。 [Determining whether environment information obtained from the sky can be used]
As shown in FIG. 7, a traveling path 601 that is an object of autonomous movement includes a region 612 in which a structure 611 such as a tree or a building exists, and a region 613 in which there is no structure around. . Note that balloons 621 and 622 from the areas 612 and 613 in FIG. 7 show bird's-eye views of the areas 612 and 613, respectively. Here, the environment information acquired by the aircraft is referred to as the sky environment information. The aircraft information described above (the travel path height information acquired from the aircraft) is information included in the sky environment information.
　図７の領域６１２のように、街路樹などで上空が覆われていたり、あるいは、トンネルなど自律移動装置１０の走行軌跡の上空が覆われている場合などにおいて、上空環境情報は使用できない。
　そこで、上空環境情報を、走行軌跡の補正に使用できるか否かを判定することが必要となってくる。
　図１５および図１６は、上空から取得した環境情報の利用可否判定の概念図を示す図である。なお、図１５、図１６では、環境情報として物体の形状データが用いられている。前記した航空機情報は、この形状データを高さ情報として利用したものである。
　図１５は自律移動環境の周辺に木や建物などの構造物６１１が存在する領域（図７の領域６１２など）に関する図であり、図１６は自律移動環境の周辺に木や建物などの構造物が存在しない領域（図７の領域６１３など）に関する図である。
　環境情報は、図１５（ａ）および図１６（ａ）に示されるように、上空から環境情報を取得する航空機８００などと、地上を走行して環境情報を取得する自律移動装置１０などから取得される。 The sky environment information cannot be used when the sky is covered with a roadside tree or the like as in the region 612 of FIG. 7 or when the sky of the traveling locus of the autonomous mobile device 10 such as a tunnel is covered.
Therefore, it is necessary to determine whether or not the sky environment information can be used for correcting the travel locus.
FIG. 15 and FIG. 16 are diagrams showing conceptual diagrams for determining whether or not to use environment information acquired from the sky. In FIGS. 15 and 16, object shape data is used as environment information. The aircraft information described above uses this shape data as height information.
FIG. 15 is a diagram related to a region where a structure 611 such as a tree or a building exists in the vicinity of the autonomous mobile environment (the region 612 in FIG. 7), and FIG. 16 illustrates a structure such as a tree or a building around the autonomous mobile environment. FIG. 8 is a diagram relating to a region in which there is no such as region 613 in FIG.
As shown in FIG. 15A and FIG. 16A, the environmental information is acquired from the aircraft 800 that acquires environmental information from the sky and the autonomous mobile device 10 that acquires environmental information while traveling on the ground. Is done.
　ここで、自律移動装置１０から得られる環境情報を地上環境情報と称することとする。
　ここで、図１５（ａ）および図１５（ｂ）に示される形状８２１は、図１５（ａ）の領域における上空環境情報から得られる形状データの断面である。同様に、図１５（ａ）および図１５（ｃ）における形状８２２は、図１５（ａ）の領域における地上環境情報から得られる形状データの断面である。
　また、図１６（ａ）および図１６（ｂ）に示される形状８３１は、図１６（ａ）の領域における上空環境情報から得られる形状データの断面である。同様に、図１６（ａ）および図１６（ｃ）における形状８３２は、図１６（ａ）の領域における地上環境情報から得られる形状データの断面である。 Here, the environment information obtained from the autonomous mobile device 10 is referred to as ground environment information.
Here, a shape 821 shown in FIGS. 15A and 15B is a cross section of shape data obtained from the sky environment information in the region of FIG. Similarly, a shape 822 in FIGS. 15A and 15C is a cross section of shape data obtained from the ground environment information in the region of FIG.
A shape 831 shown in FIGS. 16A and 16B is a cross section of shape data obtained from the sky environment information in the region of FIG. Similarly, a shape 832 in FIGS. 16A and 16C is a cross section of shape data obtained from the ground environment information in the region of FIG.
　図１５（ａ）のように、自律移動装置１０の周囲に構造物６１１が存在する場合、図１５（ｂ）および図１５（ｃ）に示すように、上空環境情報から得られる形状データ（形状８２１）と、地上環境情報から得られる形状データ（形状８２２）とは異なる形状となる。
　逆に、図１６（ａ）のように、自律移動装置１０の周囲に構造物が存在しない場合、図１６（ｂ）および図１６（ｃ）に示すように、上空環境情報から得られる形状データ（形状８３１）と、地上環境情報から得られる形状データ（形状８３２）の形状は類似する。
　このように、周囲に構造物が存在するか否かによって、上空環境情報から得られる形状データと、地上環境情報から得られる形状データとは異なるものとなる。 When the structure 611 exists around the autonomous mobile device 10 as shown in FIG. 15A, as shown in FIGS. 15B and 15C, shape data (shape) obtained from the sky environment information 821) and the shape data (shape 822) obtained from the ground environment information are different shapes.
Conversely, when there is no structure around the autonomous mobile device 10 as shown in FIG. 16 (a), shape data obtained from the sky environment information as shown in FIGS. 16 (b) and 16 (c). The shape of the shape data (shape 832) obtained from the ground environment information is similar to (shape 831).
As described above, the shape data obtained from the sky environment information differs from the shape data obtained from the ground environment information depending on whether or not a structure exists in the surroundings.
　ここで、走行軌跡補正部１０２は、上空環境情報から得られる形状データと、地上環境情報から得られる形状データとの間で、形状・模様などの対応付けを行い、類似するか否かを判定する。このようにすることで、走行軌跡補正部１０２は、自律移動装置１０で取得した移動座標における高さ情報が上空の航空機８００などから取得した高さ情報と一致（一致度が所定の値以上）するか否かを判定することができる。つまり、走行軌跡補正部１０２は、上空環境情報から得られる高さ情報に基づいて、走行軌跡（移動座標）における高さの補正が可能であるか否かを判別することができる。 Here, the traveling locus correction unit 102 associates a shape / pattern or the like between the shape data obtained from the sky environment information and the shape data obtained from the ground environment information, and determines whether or not they are similar. To do. In this way, the travel locus correction unit 102 matches the height information in the movement coordinates acquired by the autonomous mobile device 10 with the height information acquired from the aircraft 800 or the like in the sky (the degree of coincidence is a predetermined value or more). Whether or not to do so can be determined. That is, the traveling locus correction unit 102 can determine whether or not the height in the traveling locus (moving coordinates) can be corrected based on the height information obtained from the sky environment information.
　つまり、図１５に示す例では、上空環境情報から得られる形状データ（形状８２１）と、地上環境情報から得られる形状データ（形状８２２）とが一致しないため、走行軌跡補正部１０２は、上空環境情報による高さ情報で、走行軌跡（移動座標）における高さ情報の補正を行わない。
　これに対して、図１６に示す例では、上空環境情報による形状データ（形状８３１）と、地上環境情報から得られる形状データ（形状８３２）とが類似するため、走行軌跡補正部１０２は、上空環境情報による高さ情報で、走行軌跡の補正を行う。
　このように、走行軌跡補正部１０２は、航空機で取得した形状情報と、自律移動装置自身が取得した形状情報と、を比較し、航空機で取得した形状情報と、自律移動装置自身が取得した形状情報とが一致している場合、その地点の航空機で取得される高さ情報を用いて、走行軌跡の補正を行うようにすることが可能である。なお、この処理は、図５のステップＳ３０１の前に行われる。 That is, in the example shown in FIG. 15, since the shape data (shape 821) obtained from the sky environment information does not match the shape data (shape 822) obtained from the ground environment information, the traveling locus correction unit 102 The height information in the travel locus (movement coordinates) is not corrected with the height information based on the information.
On the other hand, in the example shown in FIG. 16, the shape data (shape 831) based on the sky environment information is similar to the shape data (shape 832) obtained from the ground environment information. The travel locus is corrected with the height information based on the environmental information.
As described above, the traveling locus correction unit 102 compares the shape information acquired by the aircraft with the shape information acquired by the autonomous mobile device itself, and the shape information acquired by the aircraft and the shape acquired by the autonomous mobile device itself. When the information matches, it is possible to correct the travel locus using the height information acquired by the aircraft at that point. This process is performed before step S301 in FIG.
　ここで用いられる地上環境情報による形状データと、上空環境情報による形状データとの対応付け手法はどのような方法でもよいが、例えば、走行軌跡補正部１０２は、上空環境情報と地上環境情報との形状データ間で直接類似度を計算してもよい。あるいは、走行軌跡補正部１０２は、地上環境情報のうち、路面と壁面(立体物)を分離し、周囲に存在する壁面の密度や位置の類似度から判定してもよい。また、図示しないカメラなどで撮影した画像(模様)を用いる場合、走行軌跡補正部１０２は路面模様の類似度を用いてもよい。 The method for associating the shape data based on the ground environment information and the shape data based on the sky environment information used here may be any method. For example, the traveling locus correction unit 102 may calculate the sky environment information and the ground environment information. You may calculate a similarity directly between shape data. Alternatively, the travel locus correction unit 102 may separate the road surface and the wall surface (three-dimensional object) from the ground environment information, and determine from the density of the wall surface existing in the surroundings and the similarity of the positions. In addition, when using an image (pattern) captured by a camera (not shown) or the like, the traveling locus correction unit 102 may use the degree of similarity of the road surface pattern.
　さらに、図示しないカメラなどで撮影した画像と形状データとを組み合わせる場合には、形状データのエッジ(高さ変化)と画像のエッジ(輝度変化)を対応付けることで判定するようにしてもよい。なお、使用される高さ情報がＧＰＳ座標などによる絶対座標系と同一の座標系であれば、拘束は絶対座標系としてもよく、上空環境情報で用いられている座標系と、地上権協情報で用いられている座標系とが異なる場合には、両座標系の拘束が相対座標系の拘束として補正に用いられる。
　なお、航空機情報が用いられない場合、図１５、図１６の処理は省略可能である。 Furthermore, when combining an image photographed with a camera (not shown) and the shape data, the edge (height change) of the shape data and the edge (luminance change) of the image may be associated with each other. If the height information used is the same coordinate system as the absolute coordinate system such as GPS coordinates, the constraint may be the absolute coordinate system, and the coordinate system used in the sky environment information and the terrestrial rights cooperative information When the coordinate system used in is different from that of the coordinate system, the constraint of both coordinate systems is used for correction as the constraint of the relative coordinate system.
When aircraft information is not used, the processes in FIGS. 15 and 16 can be omitted.
　以上の手順により、ＧＰＳ座標から高さ情報が十分に得られない場合などであっても、上空環境情報を得ることができれば、高さ情報の補正を含めた走行軌跡の補正が可能になる。
　また、自律移動装置１０で取得した形状データと、航空機などから取得した形状データとをマッチングさせて、航空機から取得される高さ情報の使用可否を判定することにより、走行軌跡の補正における精度を向上させることができる。
　なお、航空機８００など上空を飛行する機器から取得した形状データを用いることで、広範な領域の形状データを一度で取得することができる。 Even if the height information cannot be sufficiently obtained from the GPS coordinates by the above procedure, if the aerial environment information can be obtained, the travel locus including the correction of the height information can be corrected.
Further, by matching the shape data acquired by the autonomous mobile device 10 with the shape data acquired from an aircraft or the like, and determining whether or not the height information acquired from the aircraft can be used, the accuracy in correcting the travel locus can be improved. Can be improved.
Note that by using shape data acquired from a device flying over the aircraft 800 or the like, shape data of a wide area can be acquired at a time.
　そして、部分地図生成部１０３が、補正された部分領域の移動座標などを基に、必要に応じて、レーザ距離センサなどから取得された形状データなどを追加して、部分地図を生成する（図３のステップＳ１０３）。つまり、部分地図生成部１０３は、第２の補正を行って補正された走行軌跡を基に、地図を生成する。そして、この地図（部分地図）は、部分毎に生成される。
　ここまでの手順により、走行軌跡補正部１０２が、１回の走行で取得した３次元の走行軌跡を補正し、補正した走行軌跡などを基に部分地図を生成することができる。 Then, the partial map generation unit 103 adds shape data acquired from a laser distance sensor or the like as necessary based on the corrected movement coordinates of the partial region, and generates a partial map (see FIG. 3 step S103). That is, the partial map generation unit 103 generates a map based on the travel locus corrected by performing the second correction. And this map (partial map) is produced | generated for every part.
Through the procedure so far, the travel locus correction unit 102 can correct the three-dimensional travel locus acquired in one run and generate a partial map based on the corrected travel locus.
［部分地図の結合］
　前記したように、広域的な地図を生成するには、一回の走行で全域の環境情報（ここでは、移動座標）を取得することは現実的ではない。そこで、走行領域を分割して環境情報を取得する必要がある。分割して自律移動装置１０を走行させて取得した環境情報は、取得された時間が離れているため、以下の性質を有する。 [Combine partial maps]
As described above, in order to generate a wide-area map, it is not realistic to acquire environmental information (here, moving coordinates) for the entire area in a single run. Therefore, it is necessary to divide the travel area and acquire environmental information. The environment information acquired by dividing and traveling the autonomous mobile device 10 has the following properties because the acquired time is apart.
（１）それぞれの部分地図における移動座標は互いに拘束を有さない。つまり、各部分地図間における移動座標間がどのような位置関係にあるかがわからない状態となっている。
（２）それぞれの部分地図における移動座標取得時において、ＧＰＳ衛星の配置が大きく異なるため測位誤差の傾向が異なる。 (1) The movement coordinates in each partial map have no constraint on each other. That is, it is in a state in which it is not known what positional relationship the moving coordinates between the partial maps are.
(2) At the time of moving coordinate acquisition in each partial map, the positioning error tends to be different because the arrangement of GPS satellites is greatly different.
　このため、それぞれの部分地図における走行軌跡を補正した後に、各部分地図における移動座標を相互に補正することなく接続・統合すると、走行路における各走行軌跡などの移動座標がずれて２重になる。つまり、接続部において、移動座標のずれが存在するため、同じ地点が２重に存在することになってしまう。このようなずれは接続地点における自己位置推定の誤差となるため、部分地図間を正確に接続・統合するためには、部分地図間の移動座標の拘束を確定することが必要である。 For this reason, after correcting the travel trajectory in each partial map, if the movement coordinates in each partial map are connected and integrated without correcting each other, the travel coordinates such as each travel trajectory on the travel path are shifted and doubled. . In other words, since there is a shift in the movement coordinates at the connection portion, the same point will exist twice. Since such a shift causes an error in self-position estimation at the connection point, in order to accurately connect and integrate the partial maps, it is necessary to determine the constraint of the movement coordinates between the partial maps.
　そこで、本実施形態では、図１７および図１８に示す手法によって部分地図間の移動座標の拘束を特定する。 Therefore, in this embodiment, the movement coordinate constraint between the partial maps is specified by the method shown in FIGS. 17 and 18.
　図１７は、本実施形態に係る部分地図の接続の手順を示すフローチャートである。なお、この処理は、図３のステップＳ１０４の詳細を示すものである。各処理の詳細は、図１８を参照して後記する。
　まず、地図生成部１０４は、部分地図間において、重複している領域である地図間重複領域における移動座標の対応付けを行う（Ｓ４０１）。つまり、各部分地図には、隣り合う前記地図間において重複する領域である地図間重複領域が設定されている。 FIG. 17 is a flowchart showing a partial map connection procedure according to this embodiment. This process shows details of step S104 in FIG. Details of each process will be described later with reference to FIG.
First, the map generation unit 104 associates movement coordinates in an overlapping area between maps that are overlapping areas between partial maps (S401). That is, each partial map is set with an inter-map overlapping area that is an overlapping area between adjacent maps.
　そして、地図生成部１０４は、ステップＳ４０１で対応付けられた移動座標を基に、処理の対象となっている各部分地図における走行軌跡を補正する第３の補正を行う（Ｓ４０２）。
　続いて、地図生成部１０４は、ステップＳ４０２で補正した走行軌跡上の移動座標を基に、部分地図を結合する（Ｓ４０３）。 Then, the map generation unit 104 performs a third correction for correcting the travel locus in each partial map to be processed based on the movement coordinates associated in step S401 (S402).
Subsequently, the map generation unit 104 combines the partial maps based on the movement coordinates on the travel locus corrected in step S402 (S403).
　以下、図１８を参照して図１７に示す各処理の詳細を説明する。
　図１８は、本実施形態に係る部分地図の接続方法を説明するための図である。
　図１８に示すように、自律移動装置が走行する領域を分割した部分地図（部分領域６１０，９０１，９０２）が、部分地図間の重複領域である地図間重複領域が生じるよう、複数取得される。なお、実際には、走行路６０１全体が覆われるように部分地図（部分領域）が取得される。
　自律移動の対象とする走行路６０１において、１回目の走行で取得された部分領域６１０の環境情報（移動座標を含む）や、２回目の走行で取得された部分領域９０１の環境情報や、３回目の走行で取得された部分領域９０２の環境情報が存在するとする。このとき、各部分領域６１０，９０１，９０２の環境情報は、互いに隣接する環境情報と重複する箇所が生じるようにする。前記した走行軌跡の補正によって、各々の部分領域６１０，９０１，９０２における移動座標は、既に誤差の最大値が制限されている。従って、重複している箇所は前記した対応付けにより、相対座標系における移動座標間の拘束を得ることができる。 The details of each process shown in FIG. 17 will be described below with reference to FIG.
FIG. 18 is a diagram for explaining a partial map connection method according to the present embodiment.
As shown in FIG. 18, a plurality of partial maps ( partial regions 610, 901, 902) obtained by dividing the region where the autonomous mobile device travels are acquired so that an inter-map overlapping region that is an overlapping region between the partial maps is generated. . In practice, a partial map (partial region) is acquired so that the entire travel path 601 is covered.
In the travel route 601 to be autonomously moved, the environment information (including the movement coordinates) of the partial area 610 acquired by the first travel, the environment information of the partial area 901 acquired by the second travel, 3 It is assumed that there is environment information of the partial area 902 acquired by the second run. At this time, the environmental information of each of the partial areas 610, 901, and 902 is generated so as to overlap with the environmental information adjacent to each other. By the correction of the travel locus described above, the maximum value of the error is already limited for the movement coordinates in each of the partial areas 610, 901, and 902. Therefore, it is possible to obtain constraints between the moving coordinates in the relative coordinate system based on the above-described correspondence between the overlapping portions.
　この対応付けによって、地図生成部１０４は、部分地図間で相対座標系の拘束を満たすように、走行軌跡をさらに補正し、各部分領域６１０，９０１，９０２における接続地点にずれのない移動座標を得る。例えば、部分領域６１０での補正後の走行軌跡１１１０上に移動座標１１１１が存在し、部分領域９０１での補正後の走行軌跡１１２０上に移動座標１１２１，１１２２が存在しているものとする。地図生成部１０４は、周囲の建物などの形状データから移動座標１１１１，１１２１，１１２２の相対的な位置関係（第２の相対的な位置関係、つまり、拘束関係）を算出する。つまり、地図生成部１０４は、隣り合う地図について、地図間重複領域における環境情報同士を互いに比較することによって、地図間重複領域における相対座標間の相対的な位置関係である第２の相対的な位置関係を算出する。言い換えれば、地図生成部１０４は、ステップＳ４０１において、図５のステップＳ３０２と同様の手法で地図間重複領域における移動座標の対応付けを行う。 By this association, the map generation unit 104 further corrects the travel trajectory so as to satisfy the constraint of the relative coordinate system between the partial maps, and the moving coordinates that do not shift the connection points in the partial areas 610, 901, and 902 are obtained. obtain. For example, it is assumed that the movement coordinates 1111 exist on the corrected travel locus 1110 in the partial area 610 and the movement coordinates 1121 and 1122 exist on the corrected travel locus 1120 in the partial area 901. The map generation unit 104 calculates a relative positional relationship (second relative positional relationship, that is, a constraint relationship) of the movement coordinates 1111, 1121, and 1122 from shape data such as surrounding buildings. That is, the map generation unit 104 compares the environment information in the inter-map overlap area with each other for adjacent maps, thereby obtaining a second relative position that is a relative positional relationship between the relative coordinates in the inter-map overlap area. Calculate the positional relationship. In other words, in step S401, the map generation unit 104 associates the movement coordinates in the inter-map overlap area using the same method as in step S302 in FIG.
　この移動座標の位置関係の対応付けが、図１７のステップＳ４０１の処理に相当する。このように、地図生成部１０４は、地図間重複領域において、前記周囲の環境情報の比較によって、前記地図間重複領域を共有している部分地図における走行軌跡上の異なる地点間の位置関係を確定する。
　このようにすることで、地図間重複領域における座標間の拘束（位置関係）を特定することができるので、部分地図を人手を介さずに、かつ、正確に結合することができる。 The association of the positional relationship of the movement coordinates corresponds to the process of step S401 in FIG. As described above, the map generation unit 104 determines the positional relationship between different points on the travel locus in the partial map sharing the map overlap region by comparing the surrounding environment information in the map overlap region. To do.
By doing so, it is possible to specify the constraint (positional relationship) between the coordinates in the inter-map overlap region, so that the partial maps can be accurately combined without human intervention.
　そして、地図生成部１０４は、移動座標１１１１，１１２１，１１２２の地図上での位置関係が、確定された位置関係となるよう走行軌跡を補正（第３の補正）する。つまり、地図生成部１０４は、当該地図間重複領域における各相対座標が、算出した第２の相対的な位置関係となるよう、隣り合う地図における一方または双方の走行軌跡を変形させることで、地図における走行軌跡を補正する第３の補正を行う。
　この走行軌跡の補正が、図１７のステップＳ４０２に相当する。つまり、地図生成部１０４は、部分地図における各地点が前記確定した位置関係となるよう、処理対象となっている部分地図における走行軌跡を補正する第３の補正を行う。
　このとき、地図生成部１０４は、走行軌跡を並進させたり、回転させたりして、互いの走行軌跡上における移動座標が確定された位置関係となるようにする。 Then, the map generation unit 104 corrects the travel locus so that the positional relationship of the movement coordinates 1111, 1121, and 1122 on the map becomes the determined positional relationship (third correction). In other words, the map generation unit 104 deforms one or both traveling trajectories in adjacent maps so that each relative coordinate in the inter-map overlap area has the calculated second relative positional relationship, thereby generating a map. A third correction for correcting the travel locus at is performed.
This correction of the travel locus corresponds to step S402 in FIG. That is, the map generation unit 104 performs a third correction for correcting the travel locus on the partial map that is the processing target so that the points on the partial map have the determined positional relationship.
At this time, the map generation unit 104 translates or rotates the travel trajectory so that the movement coordinates on the travel trajectories are in a determined positional relationship.
　しかしながら、単に並進させたり、回転させたりするだけでは、ある箇所については位置関係が一致するものの、他の箇所ではずれが生じてしまうということが起こりうる。
　このような場合、地図生成部１０４は、各移動座標までの距離・道のりに応じて拘束における影響を調整してもよい。つまり、道のりや各移動座標間の量に応じるように、つまり、道のりや各移動座標間の量毎に重み付けして調整したりした上で補正を行ってもよい。
　あるいは、地図生成部１０４は、部分地図間における移動座標の位置関係（拘束関係）の重み付けを大きくした上で、図５のステップＳ３０１～Ｓ３０３の処理を行い、再度、走行軌跡の補正を行うようにしてもよい。このようにすることで、部分地図の接続における確実性を向上させることができる。 However, simply translating or rotating may cause the positional relationship to coincide with a certain part, but a shift may occur in another part.
In such a case, the map generation unit 104 may adjust the influence on the constraint according to the distance / distance to each moving coordinate. In other words, the correction may be performed in accordance with the amount of distance between the travel coordinates and the movement coordinates, that is, after adjusting by weighting the distance between the distance of travel and the movement coordinates.
Alternatively, the map generation unit 104 increases the weight of the positional relationship (constraint relationship) of the movement coordinates between the partial maps, and then performs the processing of steps S301 to S303 in FIG. 5 to correct the travel locus again. It may be. By doing in this way, the certainty in the connection of a partial map can be improved.
　このように、部分地図間で相対座標系の拘束を満たすように、走行軌跡をさらに補正することで、部分地図の接続の確実性を向上させることができる。 Thus, the certainty of the connection of the partial maps can be improved by further correcting the traveling trajectory so as to satisfy the constraint of the relative coordinate system between the partial maps.
　そして、地図生成部１０４は、ステップＳ４０２で補正した走行軌跡上の移動座標を基に、部分地図の結合を行う（図１７のステップＳ４０３）。つまり、地図生成部１０４は、第３の補正によって補正された走行軌跡における相対座標を地図間重複領域で結合することで、地図を結合する。 The map generation unit 104 then combines the partial maps based on the movement coordinates on the travel locus corrected in step S402 (step S403 in FIG. 17). In other words, the map generation unit 104 combines the maps by combining the relative coordinates in the travel locus corrected by the third correction in the inter-map overlap area.
　以上のように、図５の処理によって移動座標を予め補正しておくことで、複数の部分環境情報（特に移動座標）を接続する際に、接続箇所における環境情報の対応付けに基づく相対座標系での拘束を容易に決定することができる。このようにすることによって、広域的な走行領域であっても人手を介さずに部分地図の接続をすることができる。 As described above, when the movement coordinates are corrected in advance by the processing of FIG. 5, when connecting a plurality of pieces of partial environment information (particularly movement coordinates), a relative coordinate system based on the association of the environment information at the connection location. The constraint at can be easily determined. By doing in this way, even if it is a wide driving | running | working area | region, a partial map can be connected without a manual intervention.
　最後に、地図生成部１０４は、必要に応じて得られた各々の移動座標の周辺に、建物などの形状データなどを配置していくことで自己位置推定用の地図を完成する。このようにして、完成した地図は、各地点の路面高度の情報（高さ情報）を含んでいる。なお、同一地点であっても、誤差などが原因で、各々の移動座標の高さが完全には一致しない場合、地図生成部１０４は、近接する値から平均値を計算して任意地点における高さと見なしてもよい。 Finally, the map generation unit 104 completes a map for self-position estimation by arranging shape data such as buildings around each moving coordinate obtained as necessary. In this way, the completed map includes road surface height information (height information) at each point. Note that, even if they are at the same point, if the heights of the respective movement coordinates do not completely match due to errors or the like, the map generation unit 104 calculates an average value from adjacent values and calculates the height at an arbitrary point. You may consider that.
　また、自律移動時において取得される環境情報と、手動走行時に取得される環境情報は原則として同じであるため、自律移動させた際の移動座標を、前記した手法で補正し、補正された走行軌跡上の移動座標に基づいて地図を修正してもよい。このようにすることで、地図の部分修正を行うことができる。また、部分地図同士の境界付近の移動座標のみを補正するようにしてもよい。
　さらに、自律移動の対象とする走行路６０１（図７）で何度も走行して環境情報を取得することで、拘束が増えるため、各誤差のばらつきを平均化によって小さくすることができる。このようにすることで、地図を高精度化することができる。つまり、自律移動時に自律移動装置１０が何度も同じところを走行することによって得られる移動座標間の拘束を、走行軌跡補正部１０２が図５のステップＳ３０１～Ｓ３０３の処理を用いて確定することにより、より精度の高い補正後の走行軌跡を得ることができる。なお、移動座標などの環境情報が増え過ぎた場合には間引きや圧縮によって削減してもよく、複数回環境情報を取得した際に存在確率の低い情報については移動体などのノイズであるものとして優先的に削除するようにしてもよい。 In addition, since the environment information acquired during autonomous movement and the environment information acquired during manual driving are the same in principle, the movement coordinates when moving autonomously are corrected by the above-described method, and the corrected driving is performed. You may correct a map based on the movement coordinate on a locus | trajectory. In this way, partial map correction can be performed. Moreover, you may make it correct | amend only the movement coordinate of the boundary vicinity of partial maps.
Furthermore, since the environment is acquired by traveling many times on the travel path 601 (FIG. 7) that is the target of autonomous movement, constraints are increased, so that variation in each error can be reduced by averaging. By doing in this way, a map can be highly accurate. That is, the travel locus correction unit 102 determines the constraints between the movement coordinates obtained by the autonomous mobile device 10 traveling the same place many times during the autonomous movement, using the processing of steps S301 to S303 in FIG. Thus, it is possible to obtain a corrected travel locus with higher accuracy. In addition, if environment information such as moving coordinates increases too much, it may be reduced by thinning or compression, and information with low existence probability when acquiring environment information multiple times is assumed to be noise such as moving objects You may make it delete preferentially.
　また、本実施形態では、走行軌跡の補正を、移動座標と、ＧＰＳ座標に基づいて行っているが、これに限らず、他の自律移動装置１０が計測した環境情報などを用いてもよい。
　また、本実施形態では、走行軌跡における高さ情報の補正に、ＧＰＳと航空機による環境情報を用いているが、これに限らず、他の自律移動装置１０が計測した環境情報などを用いてもよい。 Moreover, in this embodiment, although correction | amendment of a driving | running | working locus is performed based on a movement coordinate and a GPS coordinate, you may use not only this but the environmental information etc. which the other autonomous mobile device 10 measured.
Further, in the present embodiment, environmental information by GPS and aircraft is used to correct the height information in the travel locus. However, the present invention is not limited to this, and environmental information measured by other autonomous mobile devices 10 may be used. Good.
　なお、走行軌跡は、本実施形態のように線で表現されなくてもよい。例えば、走行軌跡が点列で表示されてもよい。 Note that the travel locus may not be represented by a line as in the present embodiment. For example, the travel locus may be displayed as a point sequence.
　なお、本発明は前記した実施形態に限定されるものではなく、様々な変形例が含まれる。例えば、前記した実施形態は本発明を分かりやすく説明するために詳細に説明したものであり、必ずしも説明したすべての構成を有するものに限定されるものではない。 In addition, this invention is not limited to above-described embodiment, Various modifications are included. For example, the above-described embodiment has been described in detail for easy understanding of the present invention, and is not necessarily limited to having all the configurations described.
　また、前記した各構成、機能、各部１０１～１０７、各記憶部１２１，１２２などは、それらの一部またはすべてを、例えば集積回路で設計することなどによりハードウェアで実現してもよい。また、図２で示すように、前記した各構成、機能などは、ＣＰＵ２０１，２１１などのプロセッサがそれぞれの機能を実現するプログラムを解釈し、実行することによりソフトウェアで実現してもよい。各機能を実現するプログラム、テーブル、ファイルなどの情報は、図２に示すようにメモリ２０２や、ＲＯＭ２１３や、ＨＤ２１５に格納すること以外に、ＳＳＤ（Solid State Drive）などの記録装置、または、ＩＣ（Integrated Circuit）カードや、ＳＤ（Secure Digital）カード、ＤＶＤ（Digital Versatile Disc）などの記録媒体に格納することができる。
　また、各実施形態において、制御線や情報線は説明上必要と考えられるものを示しており、製品上必ずしもすべての制御線や情報線を示しているとは限らない。実際には、ほとんどすべての構成が相互に接続されていると考えてよい。 Each of the above-described configurations, functions, units 101 to 107, storage units 121 and 122, etc. may be realized by hardware by designing a part or all of them, for example, with an integrated circuit. Further, as shown in FIG. 2, the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by a processor such as the CPU 201, 211. Information such as programs, tables, and files for realizing each function is stored in the memory 202, the ROM 213, and the HD 215 as shown in FIG. 2, or a recording device such as an SSD (Solid State Drive) or an IC It can be stored in a recording medium such as an (Integrated Circuit) card, an SD (Secure Digital) card, or a DVD (Digital Versatile Disc).
In each embodiment, control lines and information lines are those that are considered necessary for explanation, and not all control lines and information lines are necessarily shown on the product. In practice, it can be considered that almost all configurations are connected to each other.
　本実施形態によれば、移動座標の誤差を、ＧＰＳ座標の誤差程度に抑えることができるので、部分地図の接続の確実性を向上させることができる。
　また、図５のステップＳ３０１の処理を行って、全体的に走行軌跡を補正することで、ＧＰＳ座標や、航空機情報を取得できない地点が存在しても、本来連続的な走行軌跡が不連続的になってしまうのを防ぐことができる。 According to the present embodiment, the error of the movement coordinate can be suppressed to the extent of the error of the GPS coordinate, so that the certainty of the partial map connection can be improved.
Further, by performing the process of step S301 in FIG. 5 and correcting the overall travel locus, even if there is a point where GPS coordinates or aircraft information cannot be obtained, the originally continuous travel locus is discontinuous. Can be prevented.
　本実施形態の図５のステップＳ３０１の処理により、重複走行領域の検出が可能となり、ステップＳ３０２によって重複走行領域における移動座標間の拘束（相対的な位置関係）を特定することができる。
　また、重複して走行した箇所の移動座標間の相対的な位置関係を算出することで、移動座標を対応付けた上で、第２の補正を行うことにより、移動座標の誤差をより精度高く補正することができる。
　また、図５のステップＳ３０１～Ｓ３０３の処理を行うことで、部分地図間の地図間重複領域における移動座標を対応付けを、ユーザの手を介することなく行うことができる。 The process of step S301 in FIG. 5 of the present embodiment makes it possible to detect the overlapping traveling area, and it is possible to specify the constraint (relative positional relationship) between the movement coordinates in the overlapping traveling area in step S302.
In addition, by calculating the relative positional relationship between the movement coordinates of the portions that have traveled in duplicate, the second correction is performed after associating the movement coordinates, so that the error of the movement coordinates can be made more accurate. It can be corrected.
Further, by performing the processing of steps S301 to S303 in FIG. 5, it is possible to associate the movement coordinates in the inter-map overlap area between the partial maps without the user's hand.
　本実施形態で補正された走行軌跡を用いて部分地図の生成を行い、この部分地図の結合を行うことで、精度の高い結合が可能となり、高精度の広域的な地図を生成することができる。
　移動座標が高さ情報を含むことで、高さ方向を含めた走行軌跡の補正が可能となる。
　また、移動座標がデットレコニングや、オドメトリによるものであることで、自律移動装置１０の視点による走行軌跡の補正を行うことができる。
　そして、絶対座標系に一般的な緯度・経度を用いることで、本実施形態の実現を容易にすることができる。
　さらに、高さ情報として航空機から得られる環境情報を用いることで、容易に高さ情報を得ることができる。
　さらに、自律移動装置１０で取得した形状データと、航空機などから取得した形状データとをマッチングさせて、航空機から取得される高さ情報の使用可否を判定することにより、高さ情報の補正における精度を向上させることができる。 By generating a partial map using the travel locus corrected in this embodiment and combining the partial maps, it is possible to combine with high accuracy and generate a highly accurate wide area map. .
By including the height information in the movement coordinates, it is possible to correct the travel locus including the height direction.
In addition, since the movement coordinates are based on dead reckoning or odometry, the travel locus from the viewpoint of the autonomous mobile device 10 can be corrected.
And realization of this embodiment can be made easy by using general latitude and longitude for an absolute coordinate system.
Furthermore, height information can be easily obtained by using environmental information obtained from an aircraft as height information.
Furthermore, the accuracy in correcting the height information is determined by matching the shape data acquired by the autonomous mobile device 10 with the shape data acquired from the aircraft or the like to determine whether or not the height information acquired from the aircraft can be used. Can be improved.
　１　　　自律移動システム
　１０　　自律移動装置
　２０　　管制装置
　１０１　環境情報取得部（環境情報取得手段）
　１０２　走行軌跡補正部（走行軌跡補正手段）
　１０３　部分地図生成部（地図生成手段）
　１０４　地図生成部（地図生成手段）
　１０５　自己位置推定部
　１０６　経路生成部
　１０７　移動制御部
　１２１　環境情報記憶部
　１２２　地図記憶部 DESCRIPTION OF SYMBOLS 1 Autonomous mobile system 10 Autonomous mobile device 20 Control apparatus 101 Environmental information acquisition part (environment information acquisition means)
102 Traveling locus correction unit (traveling locus correction means)
103 Partial map generator (map generator)
104 Map generator (map generator)
105 Self-position estimation unit 106 Route generation unit 107 Movement control unit 121 Environmental information storage unit 122 Map storage unit

Claims (11)

  1. 　自律移動装置の移動中における自身の相対的な位置を示す相対座標と、移動中における自身の絶対的な位置を示す絶対座標と、移動中における自身の周囲の環境情報と、を対応づけて取得する環境情報取得手段と、
    　前記絶対座標に基づく第１の誤差評価値を算出し、前記相対座標により得られる前記自律移動装置の走行軌跡を、前記第１の誤差評価値に基づいて変形させることで、前記走行軌跡を補正する第１の補正を行い、
    　前記第１の補正による補正が行われた走行軌跡を構成する軌跡同士が近接している領域を検出することによって、前記自律移動装置が重複して走行した領域である重複走行領域の検出を行い、
    　前記重複走行領域における走行軌跡上の相対座標を複数選択し、選択された各相対座標に対応付けられている前記環境情報同士を互いに比較することによって、前記選択された各相対座標間における第１の相対的な位置関係を算出し、
    　前記算出した第１の相対的な位置関係と、前記絶対座標と、の両者に基づく第２の誤差評価値を算出し、前記走行軌跡を、当該第２の誤差評価値に基づいて変形させることで、前記走行軌跡を補正する第２の補正を行う走行軌跡補正手段と、
    　を有することを特徴とする自律移動システム。     Obtained by associating relative coordinates indicating the relative position of the autonomous mobile device during movement, absolute coordinates indicating the absolute position of the autonomous mobile device during movement, and environmental information of the surrounding area during movement Environmental information acquisition means,
    A first error evaluation value based on the absolute coordinates is calculated, and the traveling locus of the autonomous mobile device obtained from the relative coordinates is deformed based on the first error evaluation value, thereby correcting the traveling locus. Perform the first correction,
    By detecting a region in which the trajectories constituting the travel trajectory corrected by the first correction are close to each other, an overlapping travel region that is a region where the autonomous mobile device has traveled in duplicate is detected. ,
    By selecting a plurality of relative coordinates on the traveling locus in the overlapping traveling area and comparing the environment information associated with each selected relative coordinate with each other, a first between the selected relative coordinates is obtained. To calculate the relative position of
    Calculating a second error evaluation value based on both the calculated first relative positional relationship and the absolute coordinate, and deforming the travel locus based on the second error evaluation value; A travel locus correction means for performing a second correction for correcting the travel locus;
    An autonomous mobile system characterized by comprising:
  2. 　前記第２の補正を行って補正された走行軌跡を基に、地図を生成する地図生成手段
    　をさらに有することを特徴とする請求の範囲第１項に記載の自律移動システム。     The autonomous mobile system according to claim 1, further comprising map generating means for generating a map based on the travel locus corrected by performing the second correction.
  3. 　前記地図は、部分ごとに生成され、
    　隣り合う前記地図間において重複する領域である地図間重複領域が設定されており、
    　前記地図生成手段は、
    　隣り合う前記地図について、前記地図間重複領域における前記環境情報同士を互いに比較することによって、前記地図間重複領域における前記相対座標間の相対的な位置関係である第２の相対的な位置関係を算出し、
    　当該地図間重複領域における各相対座標が、前記算出した第２の相対的な位置関係となるよう、前記隣り合う地図における一方または双方の走行軌跡を変形させることで、前記地図における走行軌跡を補正する第３の補正を行い、
    　当該第３の補正によって補正された走行軌跡における前記相対座標を前記地図間重複領域で結合することで、前記地図を結合する
    　ことを特徴とする請求の範囲第２項に記載の自律移動システム。     The map is generated for each part,
    An overlapping area between maps that is an overlapping area between the adjacent maps is set,
    The map generation means includes:
    For the adjacent maps, by comparing the environmental information in the inter-map overlapping region with each other, a second relative positional relationship that is a relative positional relationship between the relative coordinates in the inter-map overlapping region is obtained. Calculate
    The travel locus on the map is corrected by deforming one or both of the travel locus on the adjacent maps so that each relative coordinate in the inter-map overlap area becomes the calculated second relative positional relationship. Perform the third correction,
    The autonomous mobile system according to claim 2, wherein the maps are combined by combining the relative coordinates in the travel locus corrected by the third correction in the inter-map overlap region.
  4. 　前記相対座標および絶対座標は、高さ情報を含む
    　ことを特徴とする請求の範囲第１項乃至第３項に記載の自律移動システム。     The autonomous movement system according to any one of claims 1 to 3, wherein the relative coordinates and absolute coordinates include height information.
  5. 　前記相対座標は、前記自律移動装置で取得されたデットレコニングもしくはオドメトリによる移動座標である
    　ことを特徴とする請求の範囲第１項乃至第３項に記載の自律移動システム。     The autonomous movement system according to any one of claims 1 to 3, wherein the relative coordinates are movement coordinates by dead reckoning or odometry acquired by the autonomous movement device.
  6. 　前記絶対座標は、緯度・経度座標である
    　ことを特徴とする請求の範囲第１項乃至第３項に記載の自律移動システム     The autonomous movement system according to any one of claims 1 to 3, wherein the absolute coordinates are latitude and longitude coordinates.
  7. 　前記絶対座標は、航空機で取得される高さ情報である
    　ことを特徴とする請求の範囲第４項に記載の自律移動システム。     The autonomous mobile system according to claim 4, wherein the absolute coordinates are height information acquired by an aircraft.
  8. 　前記走行軌跡補正手段は、
    　航空機で取得された前記自律移動装置の周囲の環境情報と、前記自律移動装置自身が取得した自身の周囲の環境情報と、を比較し、
    　前記航空機で取得された環境情報と、前記自律移動装置自身が取得した環境情報とが一致している場合、前記航空機で取得される高さ情報を用いて、前記第１の補正および前記第２の補正を行う
    　ことを特徴とする請求の範囲第３項に記載の自律移動システム。     The travel locus correction means includes
    Compare the environmental information around the autonomous mobile device acquired with an aircraft with the environmental information around the autonomous mobile device itself acquired,
    When the environmental information acquired by the aircraft and the environmental information acquired by the autonomous mobile device itself match, the height information acquired by the aircraft is used to perform the first correction and the second The autonomous mobile system according to claim 3, wherein the correction is performed.
  9. 　自律移動装置の移動中における自身の相対的な位置を示す相対座標と、移動中における自身の絶対的な位置を示す絶対座標と、移動中における自身の周囲の環境情報と、を対応づけて取得する環境情報取得手段と、
    　前記絶対座標に基づく第１の誤差評価値を算出し、前記相対座標により得られる前記自律移動装置の走行軌跡を、前記第１の誤差評価値に基づいて変形させることで、前記走行軌跡を補正する第１の補正を行い、
    　前記第１の補正による補正が行われた走行軌跡を構成する軌跡同士が近接している領域を検出することによって、前記自律移動装置が重複して走行した領域である重複走行領域の検出を行い、
    　前記重複走行領域における走行軌跡上の相対座標を複数選択し、選択された各相対座標に対応付けられている前記環境情報同士を互いに比較することによって、前記選択された各相対座標間における第１の相対的な位置関係を算出し、
    　前記算出した第１の相対的な位置関係と、前記絶対座標と、の両者に基づく第２の誤差評価値を算出し、前記走行軌跡を、当該第２の誤差評価値に基づいて変形させることで、前記走行軌跡を補正する第２の補正を行う走行軌跡補正手段と、
    　を有することを特徴とする管制装置。     Obtained by associating relative coordinates indicating the relative position of the autonomous mobile device during movement, absolute coordinates indicating the absolute position of the autonomous mobile device during movement, and environmental information of the surrounding area during movement Environmental information acquisition means,
    A first error evaluation value based on the absolute coordinates is calculated, and the traveling locus of the autonomous mobile device obtained from the relative coordinates is deformed based on the first error evaluation value, thereby correcting the traveling locus. Perform the first correction,
    By detecting a region in which the trajectories constituting the travel trajectory corrected by the first correction are close to each other, an overlapping travel region that is a region where the autonomous mobile device has traveled in duplicate is detected. ,
    By selecting a plurality of relative coordinates on the traveling locus in the overlapping traveling area and comparing the environment information associated with each selected relative coordinate with each other, a first between the selected relative coordinates is obtained. To calculate the relative position of
    Calculating a second error evaluation value based on both the calculated first relative positional relationship and the absolute coordinate, and deforming the travel locus based on the second error evaluation value; A travel locus correction means for performing a second correction for correcting the travel locus;
    A control device comprising:
  10. 　前記第２の補正を行って補正された走行軌跡を基に、地図を生成する地図生成手段
    　をさらに有することを特徴とする請求の範囲第９項に記載の管制装置。     The control device according to claim 9, further comprising map generating means for generating a map based on the travel locus corrected by performing the second correction.
  11. 　前記地図は、部分ごとに生成され、
    　隣り合う前記地図間において重複する領域である地図間重複領域が設定されており、
    　前記地図生成手段は、
    　隣り合う前記地図について、前記地図間重複領域における前記環境情報同士を互いに比較することによって、前記地図間重複領域における前記相対座標間の相対的な位置関係である第２の相対的な位置関係を算出し、
    　当該地図間重複領域における各相対座標が、前記算出した第２の相対的な位置関係となるよう、前記隣り合う地図における一方または双方の走行軌跡を変形させることで、前記地図における走行軌跡を補正する第３の補正を行い、
    　当該第３の補正によって補正された走行軌跡における前記相対座標を前記地図間重複領域で結合することで、前記地図を結合する
    　ことを特徴とする請求の範囲第１０項に記載の管制装置。     The map is generated for each part,
    An overlapping area between maps that is an overlapping area between the adjacent maps is set,
    The map generation means includes:
    For the adjacent maps, by comparing the environmental information in the inter-map overlapping region with each other, a second relative positional relationship that is a relative positional relationship between the relative coordinates in the inter-map overlapping region is obtained. Calculate
    The travel locus on the map is corrected by deforming one or both of the travel locus on the adjacent maps so that each relative coordinate in the inter-map overlap area becomes the calculated second relative positional relationship. Perform the third correction,
    The control device according to claim 10, wherein the maps are combined by combining the relative coordinates in the travel locus corrected by the third correction in the overlap area between the maps.
PCT/JP2012/079970 2012-11-19 2012-11-19 Autonomous movement system and control device WO2014076844A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
JP2014546831A JP5930346B2 (en) 2012-11-19 2012-11-19 Autonomous mobile system and control device
PCT/JP2012/079970 WO2014076844A1 (en) 2012-11-19 2012-11-19 Autonomous movement system and control device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2012/079970 WO2014076844A1 (en) 2012-11-19 2012-11-19 Autonomous movement system and control device

Publications (1)

Publication Number Publication Date
WO2014076844A1 true WO2014076844A1 (en) 2014-05-22

Family

ID=50730784

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2012/079970 WO2014076844A1 (en) 2012-11-19 2012-11-19 Autonomous movement system and control device

Country Status (2)

Country Link
JP (1) JP5930346B2 (en)
WO (1) WO2014076844A1 (en)

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20170015106A (en) * 2015-07-29 2017-02-08 야마하하쓰도키 가부시키가이샤 Automatic driving vehicle
JP2017037398A (en) * 2015-08-07 2017-02-16 ヤマハ発動機株式会社 Automatic drive vehicle
JP2017191539A (en) * 2016-04-15 2017-10-19 株式会社日立産機システム Position identification system
JP2018022215A (en) * 2016-08-01 2018-02-08 村田機械株式会社 Movement teaching device and movement teaching method
KR101837821B1 (en) * 2016-02-16 2018-03-12 국방과학연구소 Method for estimating position using multi-structure filter and System thereof
CN108303075A (en) * 2017-12-29 2018-07-20 广州斯马特信息科技有限公司 Orbit generation method and system
WO2019044498A1 (en) * 2017-09-04 2019-03-07 日本電産株式会社 Mobile body, location estimation device, and computer program
JP2019049887A (en) * 2017-09-11 2019-03-28 株式会社Ihi Map generation device, map generation method and control device
KR20190088866A (en) * 2017-12-29 2019-07-29 바이두 온라인 네트웍 테크놀러지 (베이징) 캄파니 리미티드 Method, apparatus and computer readable medium for adjusting point cloud data collection trajectory
JP2019168514A (en) * 2018-03-22 2019-10-03 株式会社豊田中央研究所 Map data creation device
CN111033426A (en) * 2017-09-04 2020-04-17 日本电产株式会社 Moving object, position estimation device, and computer program
JP2020067439A (en) * 2018-10-26 2020-04-30 富士通株式会社 System and method for estimating position of moving body
JP2021060849A (en) * 2019-10-08 2021-04-15 国立大学法人静岡大学 Autonomous mobile robot and control program for autonomous mobile robot
CN113031624A (en) * 2016-03-09 2021-06-25 洋马动力科技有限公司 Travel area determination device
WO2023139935A1 (en) * 2022-01-21 2023-07-27 日立Astemo株式会社 Computing device, own-position estimating device, and map information generating method

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
RU2722356C1 (en) * 2016-07-26 2020-05-29 Ниссан Мотор Ко., Лтд. Method for self-position estimation and self-position estimation device
KR101956447B1 (en) * 2017-04-20 2019-03-12 한국과학기술원 Method and apparatus for position estimation of unmanned vehicle based on graph structure
JP7501385B2 (en) 2021-01-21 2024-06-18 株式会社豊田自動織機 Map data creation method and device

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH09145394A (en) * 1995-11-24 1997-06-06 Nippon Telegr & Teleph Corp <Ntt> Method and apparatus for measuring traveling locus
JP2005062083A (en) * 2003-08-19 2005-03-10 Komatsu Engineering Corp Survey system having function of position error correction
JP2009053561A (en) * 2007-08-28 2009-03-12 Panasonic Electric Works Co Ltd Map creating system and map creating method for autonomous moving apparatus

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH09145394A (en) * 1995-11-24 1997-06-06 Nippon Telegr & Teleph Corp <Ntt> Method and apparatus for measuring traveling locus
JP2005062083A (en) * 2003-08-19 2005-03-10 Komatsu Engineering Corp Survey system having function of position error correction
JP2009053561A (en) * 2007-08-28 2009-03-12 Panasonic Electric Works Co Ltd Map creating system and map creating method for autonomous moving apparatus

Cited By (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR102146311B1 (en) * 2015-07-29 2020-08-20 야마하하쓰도키 가부시키가이샤 Automatic driving vehicle
KR20170015106A (en) * 2015-07-29 2017-02-08 야마하하쓰도키 가부시키가이샤 Automatic driving vehicle
TWI704437B (en) * 2015-07-29 2020-09-11 日商山葉發動機股份有限公司 Self-driving vehicle
JP2017037398A (en) * 2015-08-07 2017-02-16 ヤマハ発動機株式会社 Automatic drive vehicle
KR101837821B1 (en) * 2016-02-16 2018-03-12 국방과학연구소 Method for estimating position using multi-structure filter and System thereof
CN113031624A (en) * 2016-03-09 2021-06-25 洋马动力科技有限公司 Travel area determination device
JP2017191539A (en) * 2016-04-15 2017-10-19 株式会社日立産機システム Position identification system
JP2018022215A (en) * 2016-08-01 2018-02-08 村田機械株式会社 Movement teaching device and movement teaching method
WO2019044498A1 (en) * 2017-09-04 2019-03-07 日本電産株式会社 Mobile body, location estimation device, and computer program
JPWO2019044498A1 (en) * 2017-09-04 2020-08-27 日本電産株式会社 Moving object, position estimating device, and computer program
CN111033425A (en) * 2017-09-04 2020-04-17 日本电产株式会社 Moving object, position estimation device, and computer program
CN111033426A (en) * 2017-09-04 2020-04-17 日本电产株式会社 Moving object, position estimation device, and computer program
JP2019049887A (en) * 2017-09-11 2019-03-28 株式会社Ihi Map generation device, map generation method and control device
CN108303075A (en) * 2017-12-29 2018-07-20 广州斯马特信息科技有限公司 Orbit generation method and system
KR102068419B1 (en) * 2017-12-29 2020-01-21 바이두 온라인 네트웍 테크놀러지 (베이징) 캄파니 리미티드 Method, apparatus and computer readable medium for adjusting point cloud data collection trajectory
KR20190088866A (en) * 2017-12-29 2019-07-29 바이두 온라인 네트웍 테크놀러지 (베이징) 캄파니 리미티드 Method, apparatus and computer readable medium for adjusting point cloud data collection trajectory
US11067669B2 (en) 2017-12-29 2021-07-20 Baidu Online Network Technology (Beijing) Co., Ltd. Method and apparatus for adjusting point cloud data acquisition trajectory, and computer readable medium
JP2019168514A (en) * 2018-03-22 2019-10-03 株式会社豊田中央研究所 Map data creation device
JP2020067439A (en) * 2018-10-26 2020-04-30 富士通株式会社 System and method for estimating position of moving body
JP7326720B2 (en) 2018-10-26 2023-08-16 富士通株式会社 Mobile position estimation system and mobile position estimation method
JP2021060849A (en) * 2019-10-08 2021-04-15 国立大学法人静岡大学 Autonomous mobile robot and control program for autonomous mobile robot
JP7221839B2 (en) 2019-10-08 2023-02-14 国立大学法人静岡大学 Autonomous Mobile Robot and Control Program for Autonomous Mobile Robot
WO2023139935A1 (en) * 2022-01-21 2023-07-27 日立Astemo株式会社 Computing device, own-position estimating device, and map information generating method

Also Published As

Publication number Publication date
JP5930346B2 (en) 2016-06-08
JPWO2014076844A1 (en) 2017-01-05

Similar Documents

Publication Publication Date Title
JP5930346B2 (en) Autonomous mobile system and control device
JP6827627B2 (en) Methods and systems for generating and updating vehicle environment maps
US11243081B2 (en) Slam assisted INS
Wang et al. A simple and parallel algorithm for real-time robot localization by fusing monocular vision and odometry/AHRS sensors
JP2022106924A (en) Device and method for autonomous self-position estimation
US11512975B2 (en) Method of navigating an unmanned vehicle and system thereof
CN111417871A (en) Iterative closest point processing for integrated motion estimation using high definition maps based on lidar
CN108983781A (en) A kind of environment detection method in unmanned vehicle target acquisition system
KR102627453B1 (en) Method and device to estimate position
US11340093B2 (en) Submap geographic projections
EP3799618B1 (en) Method of navigating a vehicle and system thereof
CN113587930B (en) Indoor and outdoor navigation method and device of autonomous mobile robot based on multi-sensor fusion
CN111426320A (en) Vehicle autonomous navigation method based on image matching/inertial navigation/milemeter
Li et al. Hybrid filtering framework based robust localization for industrial vehicles
JP2022027593A (en) Positioning method and device for movable equipment, and movable equipment
CN114636414A (en) High definition city map drawing
CN115790571A (en) Simultaneous positioning and map construction method based on mutual observation of heterogeneous unmanned system
CN113607166B (en) Indoor and outdoor positioning method and device for autonomous mobile robot based on multi-sensor fusion
Gupta et al. Terrain‐based vehicle orientation estimation combining vision and inertial measurements
Hoang et al. A simplified solution to motion estimation using an omnidirectional camera and a 2-D LRF sensor
CN117234203A (en) Multi-source mileage fusion SLAM downhole navigation method
Hariz et al. High-Resolution Mobile Mapping Platform Using 15-mm Accuracy LiDAR and SPAN/TerraStar C-PRO Technologies
Wei Multi-sources fusion based vehicle localization in urban environments under a loosely coupled probabilistic framework
Nowicki et al. Laser-based localization and terrain mapping for driver assistance in a city bus
Zheng et al. Indoor localization and trajectory correction with point cloud-derived backbone map

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 12888417

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2014546831

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12888417

Country of ref document: EP

Kind code of ref document: A1