CN118092208A - In-station AGV intelligent navigation method and system based on global visual servo - Google Patents
In-station AGV intelligent navigation method and system based on global visual servo Download PDFInfo
- Publication number
- CN118092208A CN118092208A CN202410487349.6A CN202410487349A CN118092208A CN 118092208 A CN118092208 A CN 118092208A CN 202410487349 A CN202410487349 A CN 202410487349A CN 118092208 A CN118092208 A CN 118092208A
- Authority
- CN
- China
- Prior art keywords
- agv
- path
- speed
- station
- target
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 47
- 230000000007 visual effect Effects 0.000 title claims abstract description 29
- 230000008859 change Effects 0.000 claims abstract description 29
- 238000007599 discharging Methods 0.000 claims abstract description 14
- 230000006870 function Effects 0.000 claims description 40
- 239000003550 marker Substances 0.000 claims description 15
- 239000006185 dispersion Substances 0.000 claims description 5
- 239000011159 matrix material Substances 0.000 claims description 5
- 239000000463 material Substances 0.000 claims description 2
- 230000009466 transformation Effects 0.000 claims description 2
- 238000004519 manufacturing process Methods 0.000 abstract description 4
- 230000000694 effects Effects 0.000 abstract description 2
- 230000008685 targeting Effects 0.000 abstract 1
- 238000004364 calculation method Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000011160 research Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000035772 mutation Effects 0.000 description 2
- 230000008447 perception Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012885 constant function Methods 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 210000001503 joint Anatomy 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
Landscapes
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention relates to the field of intelligent manufacturing, and discloses an intelligent navigation method and system for an AGV in a station based on global visual servo, which are used for identifying and targeting a station environment based on global visual, and generating a target path of the AGV by combining a path function according to the obtained station entrance positioning, station exit positioning, feeding and discharging area positioning and AGV positioning; establishing an AGV motion model, obtaining the current gesture of the AGV according to global vision, and obtaining the gesture deviation of the AGV based on the AGV motion model and an AGV target path; the obtained pose deviation of the AGV and the direction change rate of the path point of the planned pathAnd inputting a PB-PID controller to control the AGV to move, so that the AGV tracks the target path, and the intelligent navigation of the AGV is completed. By the method and the device, overshoot and yaw of the system can be prevented at any moment of path tracking, and an accurate path tracking effect is achieved.
Description
Technical Field
The invention relates to the field of intelligent manufacturing, in particular to an in-station AGV intelligent navigation method and system based on global visual servoing.
Background
The unmanned feeding and discharging system of the intelligent station is a key technology for realizing intelligent manufacturing. However, currently, AGVs (Automated Guided Vehicle: automated guided vehicles) still transport the workpiece trays along a fixed path to the tray supports of the upper and lower loading areas. The transportation path needs to be formulated by a professional and cannot be easily adjusted. Under the condition, in order to realize the autonomous positioning of the AGV, the physical marks on the fixed path are usually required to be identified and checked by means of sensors such as an on-vehicle camera, inertial navigation and the like. However, the navigation method relying on the pre-marking cannot sense the navigation environment in the whole station in real time, and cannot intelligently solve the problem of uncertainty of station layout. And, AGV is when last will transport the tray butt joint to fixed tray support, because unable perception own positioning error, need prevent error's production through accurate alignment locating pin to ensure that the arm can successfully snatch the work piece on the tray according to the coordinate that designates in advance. This reliance on a mechanical structure on an excessively high alignment approach results in high complexity and low reliability of the positioning system between the tray support and the tray.
In recent years, some researches attempt to perform station environment sensing through global vision above a station and guide multiple robots to perform cooperative feeding and discharging based on the station environment sensing. The method provides a solution with high cost performance for the problems of low intellectualization of AGV station navigation and high complexity of a tray alignment positioning system. However, based on the scheme, the AGV senses the motion environment through global vision, and errors and time delay of image processing can be introduced in the control process, so that challenges are presented to a motion control method of the AGV. In the existing research, the PID controller is widely applied to the motion control process of the AGV due to the advantages of simple structure and strong adaptability. Li Weiwei and the like, in order to overcome the defect that the robustness of the PID controller in the field of easy parameter mutation is low, the suppression of AGV track tracking error and angular velocity mutation is realized by introducing a visual neuron control algorithm and a brushless DC motor mathematical model. He Jieming and the like, in order to improve the precision and the anti-interference capability of the AGV path tracking of the forklift, a speed self-adaptive motion control method based on PID is provided by fusing a pure tracking algorithm and a Stein Li Suanfa. Li Yanwen and the like, in order to optimize the response speed and parameter online setting of the PID controller, improve the stability of AGV path tracking, and propose a path tracking method based on fuzzy PID through the fuzzification of distance deviation and angle deviation. From these studies, the current optimization of the AGV controller based on PID is mainly focused on the improvement of a system model, a fuzzy rule and the like, so that the possible overshoot and oscillation of the control system are restrained, and the improvement of the accuracy and stability of the control model is realized. However, this strategy of over-relying on accurate tuning of the PID parameters ignores the preprocessing of the input information, resulting in a loss of the control target coarse control loop. If directly applied to path navigation of a visual AGV, a control model thereof is difficult to overcome strong interference caused by high time lag only through parameter adjustment.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides an intelligent navigation method of an AGV in a station based on global visual servo, which comprises the following steps:
Step one, recognizing a station environment and positioning a target based on global vision, and adopting specific shape marker contour matching to obtain station inlet positioning, station outlet positioning and feeding and discharging area positioning; redundant positioning based on targets and markers is adopted to obtain AGV positioning;
step two, generating a target path of the AGV according to the obtained station entrance positioning, station exit positioning, loading and unloading area positioning and AGV positioning and combining a path function;
Thirdly, establishing an AGV motion model, obtaining the current gesture of the AGV according to global vision, and obtaining the gesture deviation of the AGV based on the AGV motion model and an AGV target path;
Fourthly, the obtained pose deviation of the AGV and the direction change rate of the path point of the planned path are obtained And inputting a PB-PID controller to control the AGV to move, so that the AGV tracks the target path, and the intelligent navigation of the AGV is completed.
Further, the adoption of redundant positioning based on targets and markers to obtain AGV positioning includes:
The method comprises the steps of acquiring the outline center of the AGV, and determining the rotation center of the AGV according to the outline center of the AGV and the respective centers of a pair of markers oppositely arranged on the AGV, namely the position of the AGV, and the direction of the AGV is the advancing direction.
Further, the step of generating a target path of the AGV according to the obtained station entrance positioning and exit positioning, loading and unloading area positioning and AGV positioning and combining a path function comprises the following steps:
The target path comprises a station inlet to an inlet section arc line M 1 A of a loading and unloading area, a loading and unloading area inlet to an outlet section AB of the loading and unloading area and an outlet section arc line B M 2 of the loading and unloading area to a station outlet section, wherein A is the inlet of the loading and unloading area, and B is the outlet of the loading and unloading area;
The path M 1A、BM2 is constructed by a cubic interpolation polynomial curve, the mathematical expression of which is:
Wherein the method comprises the steps of ；/>For the ordinate of the obtained path point, x is the abscissa of the path point;
station inlet to upper and lower material area inlet arc segment M 1 A:
M 1 point coordinates are The direction is/>; The point A coordinates are/>The direction is/>; Arc M 1 a satisfies the following condition, the end point function value is the same:
The end derivative values are the same:
Then:
Wherein, ；
Dispersing the obtained path into path points, wherein the distances between adjacent path points are equal, taking the length of each section of curve after dispersion as D, and the path points are expressed as:
Wherein n=1, 2, …, i, i is the number of point a, wherein Is the path point n,/>Is the abscissa of the path point n,/>Is the ordinate of the path point n,/>For the direction angle of the path point n, a relation between the path length L (x) and the abscissa x is established:
Wherein, The following equation is established:
wherein D is the length of each section of curve after the dispersion, the path length at the path point n is the sequence number of the path point;
According to And a path function, obtaining the ordinate of the path point, wherein the direction of the path point is as follows:
Wherein, Is a function in the C language, returns an azimuth; /(I)，/>The abscissa and the ordinate of the n-1 th path point are respectively;
The arc section B M 2 from the outlet of the feeding and discharging area to the outlet of the station can be obtained in the same way.
Further, the establishing the AGV motion model includes:
At any moment, the target pose of the AGV is The current pose is/>The left and right wheel speeds of the AGV are:
Wherein the method comprises the steps of The function is:
wherein L is the width of the vehicle of the AGV, u is a fixed value, For the time required for the AGV to move from the current pose to the target pose,/>Is the abscissa of the target pose of the AGV,/>Is the ordinate of the target pose of the AGV,/>The direction of the target pose of the AGV; /(I)Is the abscissa of the current pose of the AGV,/>Is the ordinate of the current pose of the AGV,/>Is the direction of the current pose of the AGV,/>For the right wheel speed of AGV,/>Is the left wheel speed of the AGV.
Further, according to global visual feedback obtain AGV current gesture, based on AGV motion model and AGV target path, obtain the position appearance deviation of AGV, include:
The target speed of AGV is Wherein/>For the target straight travel speed of AGV,/>Target rotational speed for the AGV:
the current speed of AGV is AGV pose deviation is/>：
The path deviation value dev is:
Wherein, ，/>For the current straight travel speed of AGV,/>For the current rotational speed of AGV,/>For the path deviation of AGV,/>Is the attitude deviation of the AGV.
Further, the position deviation of the AGV and the change rate of the direction of the path point of the planned path are obtainedThe PB-PID controller is input, and the AGV motion is controlled, so that the AGV tracks a target path, and the intelligent navigation of the AGV is completed, and the method comprises the following steps:
the PB-PID controller comprises a PB speed distributor and a PID controller, wherein the PB speed distributor comprises a speed generator and a speed corrector; the speed generator obtains the target speed of the AGV:
Wherein, Is the coordinates of the nth path point;
The input of the speed corrector is the change rate of the direction of the path point The output is the speed after weighted transformation; rate of change in direction of nth waypoint/>The method comprises the following steps:
Therein, wherein Is the direction of the nth path point,/>The direction of the n+1th waypoint;
weighted change rate of waypoint direction The method comprises the following steps:
Wherein, And/>Is the standard deviation of the prescription change rate of the path point,/>A change rate in the direction of the (n+1) th waypoint;
For a pair of And/>The basis functions of the weighting functions are weighted as follows:
Wherein, And/>Is the weight value of the straight speed and the rotation speed of the nth path point; And/> Coefficients that are basis functions;
the output matrix of the weight function is:
wherein diag represents a diagonal matrix;
the PB speed distributor calculates to obtain a corrected target speed of the AGV:
the input of the PID controller is obtained by the PB speed distributor And/>Feedback is/>And/>Output is control AGV/>And/>，/>For straight-going speed,/>For rotational speed, the PID controller outputs:
Wherein T is the duration of one control period, Output value for straight speed,/>Output value for rotation speed,/>Is a straight-going speed proportional parameter,/>Is a straight-going speed differential parameter,/>Is a rotation speed proportional parameter,/>Is a rotational speed differential parameter.
The in-station AGV intelligent navigation system based on the global visual servo is applied to the in-station AGV intelligent navigation method based on the global visual servo, and comprises a global visual module, a target positioning module, a digital logistics path module and a path tracking module;
The global vision module, the digital logistics path module and the path tracking module are respectively connected with the target positioning module; the digital logistics path module is connected with the path tracking module.
The beneficial effects of the invention are as follows: the technical scheme provided by the invention starts from the research requirement of AGV trackless navigation in the loading and unloading stations, and utilizes the advantage that the global vision can sense the path change rate in advance, thereby providing a global vision servo control model taking a PID controller based on the path as a core. The use of global vision improves the perception capability of the navigation system, and can acquire more extensive and comprehensive environmental information, so that the feeding and discharging digital logistics paths can be generated in real time according to the layout of stations. Therefore, the PB-PID controller provided by the invention can also preprocess the input of the controller according to the path change, and ensure that overshoot and yaw of the system can be prevented at any moment of path tracking, thereby achieving an accurate path tracking effect.
Drawings
FIG. 1 is a flow chart of an intelligent navigation method of an AGV in a station based on global visual servoing;
FIG. 2 is a global visual servo based AGV motion control model;
FIG. 3 is a schematic representation of a design of a marker;
FIG. 4 is a schematic view of tray support positioning;
FIG. 5 is a schematic view of AGV positioning;
FIG. 6 is a schematic diagram of a digitized flow path;
FIG. 7 is a schematic view of an AGV motion model;
FIG. 8 is a schematic diagram of a IBGVS motion control model;
FIG. 9 is a schematic diagram of a PB-PID controller.
Detailed Description
The technical solution of the present invention will be described in further detail with reference to the accompanying drawings, but the scope of the present invention is not limited to the following description.
The features and capabilities of the present invention are described in further detail below in connection with the examples.
As shown in fig. 1, specifically, the present study proposes an AGV motion control method based on global vision, in which, first, the station environment is identified and targeted based on global vision. The inlet and the outlet of the station are positioned by the markers, and the loading and unloading areas are positioned by the workpiece tray support. The pose of the above target is used to generate the logistic path of the AGV. The position of the AGV is obtained through the auxiliary positioning of the outer contour and the marker. And finally, jointly taking the pose deviation and the logistics path of the AGV as the input of a PB-PID controller, and constructing an AGV motion control model based on global visual servo.
As shown in FIG. 2, the identified and located targets are the pallet supports, markers, and AGVs. The globally visually derived image typically contains a large amount of garbage, and the purpose of the recognition is to screen out the object from the large amount of garbage. The purpose of the positioning is to determine the pose of the target for application in the generation and motion control of digital logistic paths.
Design of markers a set of universal markers is designed herein for digital logistic path generation and AGV control. As shown in fig. 3, the markers are classified into isosceles triangle markers and "T" type markers. The triangle mark is stuck on the AGV to assist the positioning and tracking of the AGV. The T-shaped markers are stuck to the inlet and the outlet of the station and are used for generating a digital logistics path in the station.
The principle of visual screening is contour matching. And (3) obtaining the outlines of all images under the global vision through image processing, and performing outline matching with the appearance of the AGV, the tray bracket and the appearance of the marker. Contour matching employs advanced features of the contour: the Hu moment, a constant function of a contour, has translational, rotational, and scaling invariance.
The positioning of the marker is shown in fig. 3, M is the centroid of the marker outline, and A is a point on the symmetry axis of the marker. The point A of the triangular marker is determined as the vertex of an isosceles triangle, and the point A of the T-shaped marker is determined as the center point of the minimum circumscribed rectangle. The position of the marker is determined as M point and the direction is AM.
As shown in fig. 4, S 1、S2 is the centroid of the tray support, and C is the midpoint of the line segment S 1S2. C is used as the position of the tray support and is also the center point of the feeding and discharging area. A is an inlet of the feeding and discharging area, B is an outlet of the feeding and discharging area, and the direction of the tray support is from C to B. The length of DE is equal to the length of the pallet support, and the distance between AD and EB is greater than half the length of the AGV, otherwise collision may occur.
The AGV is positioned by a redundant positioning method based on targets and markers, and the positioning method based on targets only can cause tracking failure when the surface of the AGV is covered by other objects. The redundant positioning method can ensure that the tracking of the target can still be completed only by the marker when the AGV contour is lost. As shown in fig. 5, M 1、M2 is the center of the marker, C is the center of the contour of the AGV, and R is the center of rotation of the AGV. The rotation center of the AGV is necessarily on the line or extension of the contour centroid and the marker centroid, and the proportional relationship is determined. The position of the AGV is determined as its center of rotation and the direction is determined as the direction of travel.
The role of the digitizing flow path is two: (1) The digital logistics path connects the actual station inlet and outlet with the loading and unloading area in series, the path points are generated in real time and stored in a computer, and the digital logistics path is not required to be fixedly arranged in the station so as to cope with the uncertainty of the layout of the station under various working conditions; (2) The digitized flow path is used as an input to the IBGVS controller to participate in the motion control of the AGV. The pose of the station inlet and outlet and the feeding and discharging area is determined in advance according to actual requirements. IBGVS is a global visual servo control model, namely Image Based Global Visual Servoing, hereinafter abbreviated as IBGVS.
The digitized logistics path of the AGV is shown in FIG. 6, the view of FIG. 6 being that of the global vision camera overhead station. The paths in the feeding and discharging areas are straight lines. As can be seen, since the entire digitized flow path requires a smooth connection of straight line segment AB between the two directionally demanded target points M 1 and M 2. This means that the change of the generated path curve between the target points must be continuous and the first derivative (slope) and the second derivative (curvature) are also continuous. Under this condition, the present study takes the commonly used cubic interpolation polynomial function as an example of the path function connecting two target points. The function value and the derivative value of the starting point and the ending point are considered by the cubic interpolation polynomial, so that the path at the connecting point is continuous and smooth, and the situation that the curve has abrupt change or folding points is avoided.
As shown in fig. 6, path M 1A、BM2 is constructed by a cubic interpolation polynomial curve. To simplify the calculation, the mathematical expression of the cubic interpolation polynomial is:
Take the example of generating arc M 1 a: m 1 point coordinates are The direction is/>; The point A coordinates are/>The direction is/>. The following conditions need to be satisfied, the end point function values being the same:
The end derivative values are the same:
（3）
Then:
（4）
Wherein, 。
Finally, the obtained path is discretized into path points, and the distances between adjacent path points are ensured to be equal. Taking the length of each section of curve after the dispersion as D. The information of the route point is composed of two elements of position and direction, and thus the route point can be expressed as:
Where n=1, 2, …, i, i is the number of point a. Establishing a relation between the path length L (x) and the abscissa x:
Wherein, . The "length" of a function generally refers to the arc length of the function graph, i.e., the distance of the function curve from one point to another. To calculate the arc length of a function f (x) over the intervals a, b, we can use the method of integration. For a function y=f (x) in rectangular coordinates, its arc length L can be calculated by the following formula:
wherein/> Is the derivative of the function f (x)
A, b: coordinates of the start and end points of the curve segment on the x-axis
: Is a function of the slope of the curve at any point x
When we talk about calculating the arc length of the function f (x) over a certain interval, it is actually the slope of the tangent at each point on the curve that is tried to be found, and then use this slope to calculate the straight line distance (i.e. the infinitesimal arc length) of the infinitesimal curve. The arc length of the infinite small segment can be calculated by a infinitesimal method in a calculus, and specifically using the Pythagorean theorem, the arc length can be expressed as follows:
。
taking the length of each curve segment as D, the following equation can be established to solve n:
To be found And (3) carrying the coordinate values into the path function to obtain the ordinate of the point, and obtaining the x and y coordinate values of each discrete path point. The direction of the waypoint is:
the function of this formula is to define the waypoint direction Atan2: atan2 is a function, returns in the C language refer to azimuth, and the prototype of the function of atan2 in the C language is double atan2 (double y, double x), returns the arctangent of y/x in radians. The sign of the values of y and x determines the correct quadrant. It can also be understood that the argument of the complex number x+yi is calculated, and that atan2 is stable than atan at the time of calculation. x n,yn: coordinates of an nth path point; x n-1,yn-1: coordinates of the n-1 th waypoint.
A four-wheel differential AGV is selected as a research object, and a motion model is established by using a motion model establishing method of the differential AGV, as shown in figure 7.
At any time, the target pose of the AGV may be defined asThe current pose may be defined as. The left and right wheel speeds of the AGV are obtained by:
Wherein L is the width of the vehicle of the AGV, u is a fixed value, Is the time required to move from the current pose to the target pose. The target speed of an AGV is defined herein as/>, based on the motion model：
The current speed of the AGV is defined asThe pose bias of AGVs is defined as/>：
And the overlap ratio of the AGV and the path in the path tracking process is measured by using the path deviation value dev, and the smaller the path deviation degree is, the higher the overlap ratio of the path is.
Wherein,。
IBGVS can be described simply as shown in figure 8. Comparing the current gesture of the AGV obtained through global visual feedback to obtain gesture deviation of the AGVAnd directly calculated path deviation/>. In addition to this, there is a rate of change of the direction of the waypoints/>, under global path planning. Finally will/>、/>And/>And a PB-PID controller is input, wherein the PB-PID controller is a path-based PID controller, namely Path Based PID controller, namely the PB-PID controller for short, and the deviation closed-loop control based on global vision is formed. The deviation closed-loop control based on the global vision forms IBGVS control, and the current state and pose deviation of the AGV are obtained, so that the movement of the AGV is controlled, and the AGV can accurately track the target path.
The essence of AGV control is to adjust the speed input of the AGV in real time to track the planned path. The core in the IBGVS framework is a PB-PID controller, which is divided into a PB speed distributor and a PID controller, as shown in FIG. 9. The controller takes pose deviation and path change rate as reference input, takes current straight speed and rotation speed of the AGV as feedback, and takes target straight speed and rotation speed as output. The PB speed distributor and the PID controller are controlled based on the system model, the accuracy requirement on the data model is low, and the system has good robustness. Input parameter pose deviation of controllerThe expression over time is:
Therein, wherein Is the abscissa of the current pose of the AGV at time t,/>Is the direction of the current pose of the AGV at time t.
Here, the PB speed distributor is composed of a speed generator and a speed corrector. The function of the speed generator is to calculate the deviation according to the poseThe straight travel and rotational speed are calculated. The speed corrector further corrects the straight running speed and the rotating speed according to the path change rate, so that the running overshoot and the transverse swing in the running process of the AGV are restrained.
A speed generator: the speed generator calculates the target speed of the AGV according to equation (9). Deviation of target speed and poseThe calculation method is as follows:
Wherein, Is the coordinates of the nth path point.
A speed corrector: the instability that occurs during the travel of an AGV is caused by two phenomena: travel overshoot and yaw. After the calculation of the speed generator, the system can be initially calculated according to the pose deviationAnd/>. Constant/>The travel overshoot is generated when the AGV does not decelerate during the path change. /(I)The smaller the value, the/>The larger the amplitude of the variation of (c) is,The more frequently the direction of the AGV changes, resulting in a roll of the AGV.
The input to the speed corrector is the rate of change of the waypoint directionThe output is the weighted transformed speed. Defining the rate of change of the direction of the nth waypoint/>The method comprises the following steps:
The direction change rate of all path points on the digital logistics path is calculated, and the maximum direction change rate can be obtained . Further, by its standard deviation pair/>Weighting, namely, the weighted change rate/>, of the direction of the path pointThe calculation mode of (a) is as follows:
Wherein, And/>Is the standard deviation of the prescription rate of change of the waypoint.
For a pair ofAnd/>The basis functions of the weighting functions are weighted as follows:
（16）
Then the coefficient And/>The method comprises the following steps:
Wherein, And/>Is the weight value of the straight speed and the rotation speed of the nth path point; /(I)、Respectively the maximum weight value and the minimum weight value of the straight-going speed, and are fixed values,/>、/>The maximum weight value and the minimum weight value of the rotation speed are respectively, and are fixed values; /(I)Less than zero, decreasing in the right half of the y-axis, so the minimum will be at/>The output matrix of the final weight function can be obtained by the above calculation method:
in summary, the PB speed allocator calculates the target speed after AGV correction:
PID control is a control strategy that is widely used in many systems and devices. Its main purpose is to enable the system to reach the desired output more accurately, more quickly and more stably, while improving the robustness and adaptability of the system.
The input to the controller is calculated by the PB distributorAnd/>Feedback is/>And/>The output is the actual control AGV/>And/>，/>For straight-going speed,/>Is the rotational speed.
Because of the relative sliding between the ground and the wheels, the odometer has lower positioning precision and lower self-repairing capability of positioning errors in the long-term running process. However, the positioning accuracy of the odometer is relatively high in the short-term operation process, the measured AGV speed is relatively accurate, and the odometer can be used as input of a proportional term and a differential term. Thus, the PID controller used herein is actually a PD controller, expressed as:
（19）。
The foregoing is merely a preferred embodiment of the invention, and it is to be understood that the invention is not limited to the form disclosed herein but is not to be construed as excluding other embodiments, but is capable of numerous other combinations, modifications and environments and is capable of modifications within the scope of the inventive concept, either as taught or as a matter of routine skill or knowledge in the relevant art. And that modifications and variations which do not depart from the spirit and scope of the invention are intended to be within the scope of the appended claims.
Claims (7)
1. The in-station AGV intelligent navigation method based on the global visual servo is characterized by comprising the following steps of:
Step one, recognizing a station environment and positioning a target based on global vision, and adopting specific shape marker contour matching to obtain station inlet positioning, station outlet positioning and feeding and discharging area positioning; redundant positioning based on targets and markers is adopted to obtain AGV positioning;
step two, generating a target path of the AGV according to the obtained station entrance positioning, station exit positioning, loading and unloading area positioning and AGV positioning and combining a path function;
Thirdly, establishing an AGV motion model, obtaining the current gesture of the AGV according to global vision, and obtaining the gesture deviation of the AGV based on the AGV motion model and an AGV target path;
Fourthly, the obtained pose deviation of the AGV and the direction change rate of the path point of the planned path are obtained And inputting a PB-PID controller to control the AGV to move, so that the AGV tracks the target path, and the intelligent navigation of the AGV is completed.
2. The intelligent navigation method for an in-station AGV based on global visual servoing according to claim 1, wherein said obtaining the AGV position by using redundant positioning based on the target and the marker comprises:
The method comprises the steps of acquiring the outline center of the AGV, and determining the rotation center of the AGV according to the outline center of the AGV and the respective centers of a pair of markers oppositely arranged on the AGV, namely the position of the AGV, and the direction of the AGV is the advancing direction.
3. The intelligent navigation method for the AGV in the station based on the global visual servoing according to claim 2, wherein generating the target path of the AGV according to the obtained station entrance positioning and exit positioning, the feeding and discharging area positioning and the AGV positioning by combining the path function comprises:
The target path comprises a station inlet to an inlet section arc line M 1 A of a loading and unloading area, a loading and unloading area inlet to an outlet section AB of the loading and unloading area and an outlet section arc line B M 2 of the loading and unloading area to a station outlet section, wherein A is the inlet of the loading and unloading area, and B is the outlet of the loading and unloading area;
The path M 1A、BM2 is constructed by a cubic interpolation polynomial curve, the mathematical expression of which is:
Wherein the method comprises the steps of ；/>For the ordinate of the obtained path point, x is the abscissa of the path point;
station inlet to upper and lower material area inlet arc segment M 1 A:
M 1 point coordinates are The direction is/>; The point A coordinates are/>The direction is/>; Arc M 1 a satisfies the following condition, the end point function value is the same:
The end derivative values are the same:
Then:
Wherein, ；
Dispersing the obtained path into path points, wherein the distances between adjacent path points are equal, taking the length of each section of curve after dispersion as D, and the path points are expressed as:
Wherein n=1, 2, …, i, i is the number of point a, wherein Is the path point n,/>Is the abscissa of the path point n,/>Is the ordinate of the path point n,/>For the direction angle of the path point n, a relation between the path length L (x) and the abscissa x is established:
Wherein, The following equation is established:
wherein D is the length of each section of curve after the dispersion, the path length at the path point n is the sequence number of the path point;
According to And a path function, obtaining the ordinate of the path point, wherein the direction of the path point is as follows:
Wherein, Is a function in the C language, returns an azimuth; /(I)，/>The abscissa and the ordinate of the n-1 th path point are respectively;
The arc section B M 2 from the outlet of the feeding and discharging area to the outlet of the station can be obtained in the same way.
4. The intelligent navigation method for the AGV in the station based on the global visual servoing according to claim 3, wherein said establishing the AGV motion model comprises:
At any moment, the target pose of the AGV is The current pose is/>The left and right wheel speeds of the AGV are:
Wherein the method comprises the steps of The function is:
wherein L is the width of the vehicle of the AGV, u is a fixed value, For the time required for the AGV to move from the current pose to the target pose,/>Is the abscissa of the target pose of the AGV,/>Is the ordinate of the target pose of the AGV,/>The direction of the target pose of the AGV; Is the abscissa of the current pose of the AGV,/> Is the ordinate of the current pose of the AGV,/>For the direction of the current pose of the AGV,For the right wheel speed of AGV,/>Is the left wheel speed of the AGV.
5. The intelligent navigation method for the AGV in the station based on the global visual servo according to claim 4, wherein the step of obtaining the current gesture of the AGV based on the global visual feedback and the gesture deviation of the AGV based on the AGV motion model and the AGV target path comprises the steps of:
The target speed of AGV is Wherein/>For the target straight travel speed of AGV,/>Target rotational speed for the AGV:
the current speed of AGV is AGV pose deviation is/>：
The path deviation value dev is:
Wherein, ，/>For the current straight travel speed of AGV,/>For the current rotational speed of AGV,/>For the path deviation of AGV,/>Is the attitude deviation of the AGV.
6. The intelligent navigation method of the AGV in the station based on the global visual servoing according to claim 5, wherein the pose deviation of the AGV and the change rate of the direction of the path point of the planned path are obtainedThe PB-PID controller is input, and the AGV motion is controlled, so that the AGV tracks a target path, and the intelligent navigation of the AGV is completed, and the method comprises the following steps:
the PB-PID controller comprises a PB speed distributor and a PID controller, wherein the PB speed distributor comprises a speed generator and a speed corrector; the speed generator obtains the target speed of the AGV:
Wherein, Is the coordinates of the nth path point;
The input of the speed corrector is the change rate of the direction of the path point The output is the speed after weighted transformation; rate of change in direction of nth waypoint/>The method comprises the following steps:
Therein, wherein Is the direction of the nth path point,/>The direction of the n+1th waypoint;
weighted change rate of waypoint direction The method comprises the following steps:
Wherein, And/>Is the standard deviation of the prescription change rate of the path point,/>A change rate in the direction of the (n+1) th waypoint;
For a pair of And/>The basis functions of the weighting functions are weighted as follows:
Wherein, And/>Is the weight value of the straight speed and the rotation speed of the nth path point; /(I)And/>Coefficients that are basis functions;
the output matrix of the weight function is:
wherein diag represents a diagonal matrix;
the PB speed distributor calculates to obtain a corrected target speed of the AGV:
the input of the PID controller is obtained by the PB speed distributor And/>Feedback is/>And/>The output is the control of the straight speed/>, of the AGVAnd rotational speed/>The PID controller outputs:
Wherein T is the duration of one control period, Output value for straight speed,/>In order to output a value for the rotational speed,Is a straight-going speed proportional parameter,/>Is a straight-going speed differential parameter,/>Is a rotation speed proportional parameter,/>Is a rotational speed differential parameter.
7. The in-station AGV intelligent navigation system based on the global visual servo is characterized by comprising a global visual module, a target positioning module, a digital logistics path module and a path tracking module, wherein the in-station AGV intelligent navigation method based on the global visual servo is applied to any one of claims 1-6;
The global vision module, the digital logistics path module and the path tracking module are respectively connected with the target positioning module; the digital logistics path module is connected with the path tracking module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410487349.6A CN118092208A (en) | 2024-04-23 | 2024-04-23 | In-station AGV intelligent navigation method and system based on global visual servo |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410487349.6A CN118092208A (en) | 2024-04-23 | 2024-04-23 | In-station AGV intelligent navigation method and system based on global visual servo |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118092208A true CN118092208A (en) | 2024-05-28 |
Family
ID=91157278
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410487349.6A Pending CN118092208A (en) | 2024-04-23 | 2024-04-23 | In-station AGV intelligent navigation method and system based on global visual servo |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118092208A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2014034246A (en) * | 2012-08-07 | 2014-02-24 | Honda Motor Co Ltd | Steering control device for vehicle |
| CN109579863A (en) * | 2018-12-13 | 2019-04-05 | 北京航空航天大学 | Unknown topographical navigation system and method based on image procossing |
| CN113175324A (en) * | 2021-04-27 | 2021-07-27 | 中铁工程装备集团有限公司 | Cascade composite PID control method for coupled motion of large arm of drill jumbo |
| CN116198522A (en) * | 2023-05-05 | 2023-06-02 | 江苏大学 | Unmanned mining card transverse and vertical coupling hierarchical control method for complex mining area working conditions |
| CN116512237A (en) * | 2022-11-28 | 2023-08-01 | 广东建石科技有限公司 | Industrial robot vision servo method, device, electronic equipment and storage medium |
-
2024
- 2024-04-23 CN CN202410487349.6A patent/CN118092208A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2014034246A (en) * | 2012-08-07 | 2014-02-24 | Honda Motor Co Ltd | Steering control device for vehicle |
| CN109579863A (en) * | 2018-12-13 | 2019-04-05 | 北京航空航天大学 | Unknown topographical navigation system and method based on image procossing |
| CN113175324A (en) * | 2021-04-27 | 2021-07-27 | 中铁工程装备集团有限公司 | Cascade composite PID control method for coupled motion of large arm of drill jumbo |
| CN116512237A (en) * | 2022-11-28 | 2023-08-01 | 广东建石科技有限公司 | Industrial robot vision servo method, device, electronic equipment and storage medium |
| CN116198522A (en) * | 2023-05-05 | 2023-06-02 | 江苏大学 | Unmanned mining card transverse and vertical coupling hierarchical control method for complex mining area working conditions |
Non-Patent Citations (5)
| Title |
|---|
| FARKH, RIHEM等: "Vision Navigation Based PID Control for Line Tracking Robot", INTELLIGENT AUTOMATION AND SOFT COMPUTING, vol. 35, no. 1, 7 August 2022 (2022-08-07), pages 901 - 911 * |
| 刘洁美: "基于路面附着系数预估的智能汽车纵向速度规划及控制策略研究", 中国优秀硕士学位论文全文数据库 工程科技II辑, no. 1, 31 January 2024 (2024-01-31), pages 035 - 505 * |
| 吴连伟, 等: "基于PLC控制的AGV智能物流输送***的研究与应用", 机械工程师, no. 09, 30 September 2023 (2023-09-30), pages 142 - 146 * |
| 武星: "视觉导航的轮式移动机器人运动控制技术研究", 中国博士学位论文全文数据库, no. 01, 31 January 2012 (2012-01-31), pages 140 - 30 * |
| 段玉龙, 等: "基于自适应模糊PID算法的蔗种调向装置定位控制", 控制工程, 7 June 2023 (2023-06-07), pages 1 - 9 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5073749A (en) | Mobile robot navigating method | |
| US20160107682A1 (en) | System and method for vehicle steering control | |
| CN112462760B (en) | Double-steering-wheel AGV path tracking method | |
| CN109656250A (en) | A kind of path following method of laser fork truck | |
| CN110837257A (en) | AGV composite positioning navigation system based on iGPS and vision | |
| CN114442491A (en) | Local path planning and tracking control method, device and medium for indoor robot | |
| CN111679676A (en) | AGV movement track control method | |
| Sahoo et al. | Design and development of a heading angle controller for an unmanned ground vehicle | |
| Juntao et al. | Research of AGV positioning based on the two-dimensional Code Recognition Method | |
| CN115993089B (en) | PL-ICP-based online four-steering-wheel AGV internal and external parameter calibration method | |
| TWI770966B (en) | Guidance control method of unmanned self-propelled vehicle | |
| CN118092208A (en) | In-station AGV intelligent navigation method and system based on global visual servo | |
| CN111830980B (en) | Laser navigation path following method | |
| Yan et al. | A New Path Tracking Algorithm for Four-Wheel Differential Steering Vehicle | |
| CN114089730A (en) | Robot motion planning method and automatic guided vehicle | |
| Xiao et al. | Smooth trajectory tracking using longitudinal distance constraint for a 4WS4WD unmanned ground vehicle | |
| TWI770965B (en) | Guidance control method of unmanned self-propelled vehicle | |
| Avanzini et al. | A control strategy taking advantage of inter-vehicle communication for platooning navigation in urban environment | |
| CN117519190B (en) | Novel articulated vehicle control method | |
| CN117850413B (en) | Vehicle control method based on broken line path | |
| CN113467460B (en) | Agricultural machine path tracking method and system based on double-circular forward looking distance | |
| CN114253241B (en) | Path tracking method for industrial intelligent trolley | |
| Szͩcs et al. | Experimental verification of a control system for autonomous navigation | |
| Dave et al. | Tuning procedure for correction of systematic errors in a quad configuration agv | |
| Zang et al. | Optimal path tracking controller of multiple unmanned tracked vehicles: A distributed model predictive control approach |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |