CN115373399A - Ground robot path planning method based on air-ground cooperation - Google Patents
Ground robot path planning method based on air-ground cooperation Download PDFInfo
- Publication number
- CN115373399A CN115373399A CN202211108617.6A CN202211108617A CN115373399A CN 115373399 A CN115373399 A CN 115373399A CN 202211108617 A CN202211108617 A CN 202211108617A CN 115373399 A CN115373399 A CN 115373399A
- Authority
- CN
- China
- Prior art keywords
- path
- node
- ground
- map
- robot
- 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
- 230000000694 effects Effects 0.000 claims abstract description 9
- 238000005457 optimization Methods 0.000 claims abstract description 7
- 230000006870 function Effects 0.000 claims description 39
- 230000033001 locomotion Effects 0.000 claims description 31
- 230000011218 segmentation Effects 0.000 claims description 16
- 238000012549 training Methods 0.000 claims description 16
- 230000008569 process Effects 0.000 claims description 12
- 238000009499 grossing Methods 0.000 claims description 9
- 230000008859 change Effects 0.000 claims description 8
- 238000012360 testing method Methods 0.000 claims description 6
- 230000004927 fusion Effects 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000011478 gradient descent method Methods 0.000 claims description 4
- 239000000203 mixture Substances 0.000 claims description 4
- 238000013527 convolutional neural network Methods 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 claims description 3
- 230000009191 jumping Effects 0.000 claims description 3
- 230000015572 biosynthetic process Effects 0.000 description 10
- 238000005381 potential energy Methods 0.000 description 7
- 230000004888 barrier function Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000010276 construction Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 238000006073 displacement reaction Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000010845 search algorithm Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 206010000369 Accident Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 238000004804 winding Methods 0.000 description 1
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
- G05D1/0231—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means
- G05D1/0238—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using obstacle or wall sensors
- G05D1/024—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using obstacle or wall sensors in combination with a laser
-
- 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
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0214—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with safety or protection criteria, e.g. avoiding hazardous areas
-
- 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
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0221—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving a learning process
-
- 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
- G05D1/0231—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means
- G05D1/0246—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using a video camera in combination with image processing means
- G05D1/0251—Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means using a video camera in combination with image processing means extracting 3D information from a plurality of images taken from different locations, e.g. stereo vision
-
- 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
- G05D1/0257—Control of position or course in two dimensions specially adapted to land vehicles using a radar
-
- 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
-
- 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
- G05D1/0287—Control of position or course in two dimensions specially adapted to land vehicles involving a plurality of land vehicles, e.g. fleet or convoy travelling
- G05D1/0289—Control of position or course in two dimensions specially adapted to land vehicles involving a plurality of land vehicles, e.g. fleet or convoy travelling with means for avoiding collisions between vehicles
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Aviation & Aerospace Engineering (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Electromagnetism (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Optics & Photonics (AREA)
- Multimedia (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention relates to a ground robot path planning method based on air-ground coordination, which comprises the following steps: constructing a complete space three-dimensional model of the target area based on the space-based image acquisition data of the target area; constructing a moving path area of the ground mobile robot based on the space three-dimensional model, and rasterizing the moving path area to obtain a binary grid map of a path which can be accessed by the robot; and establishing a mixed map combined with the Voronoi map based on the binary grid map, carrying out mixed A-algorithm global path planning on the mixed map, and carrying out path optimization based on the multi-guide-point traction effect to obtain a reliable moving path of the ground robot. The ground robot path planning method based on the air-ground coordination enables the path planned by the robot to be drawn to the vicinity of a plurality of set guide points, and a safe and reliable path is successfully planned under the environment of multiple obstacles.
Description
Technical Field
The invention belongs to the technical field of air-ground cooperation and path planning, and particularly relates to a ground robot path planning method based on air-ground cooperation.
Background
In recent years, air-ground collaboration is a leading-edge research hotspot problem of cooperative work of heterogeneous unmanned systems. Aerial unmanned aerial vehicle group and ground mobile robot group collaborative work compare in solitary ground robot group or aerial unmanned aerial vehicle group more have the advantage. The unmanned aerial vehicle has ultrahigh maneuvering flight capability and wide aerial visual field, can quickly provide global information for the ground robot, and the ground robot can observe the state (local information) of a target area at a short distance, and can be applied to tasks such as fire rescue, environmental exploration, target search, military reconnaissance and the like by combining the advantages of the ground robot and the ground robot.
Aiming at the reconnaissance and rescue tasks of the heterogeneous unmanned system in a large-scale fire accident field, due to the complexity of fire extinguishing environment and the requirement on the real-time property of environment perception, in order to realize that the ground multi-fire-fighting robot can independently navigate in an unknown environment, map construction and path planning are very important links in the navigation technology, and the problem of high time consumption of ground robot map construction can be effectively solved by adopting a multi-unmanned aerial vehicle cooperative map construction mode. Therefore, the multi-unmanned aerial vehicle based on the real-time three-dimensional reconstruction technology is adopted to realize rapid high-precision three-dimensional space collaborative map building for the fire scene environment, a reliable moving area map is built for the ground robot, and the binary grid map is rasterized according to the extracted reliable moving area map to obtain a grid map which can be used for path planning. At present, common two-dimensional path planning algorithms based on graph search include an algorithm a, a Dijkstra, an RRT algorithm, a D algorithm and the like, and these path planning algorithms are based on position level search and do not consider the problem of actual orientation of robot motion. If the method is directly applied to the path planning of the intelligent fire-fighting robot, problems are bound to be caused: 1) In practical application, the direction of the robot needs to be adjusted in advance before executing a path tracking task each time, and is consistent with the initial direction of a planned path as much as possible, otherwise, the speed of the robot is easy to change too much, so that potential safety hazards are caused; 2) Considering that when the fire-fighting robot drags the water hose to move, the problem that the direction of the path cannot be ensured by a two-dimensional path planning algorithm because the planned path needs to be smooth enough and the direction of the path is consistent with the direction of the robot as much as possible because the robot is prevented from rotating or retreating to cause the water hose to be bent and burst due to large amplitude is solved; 3) If the group fire-fighting robot takes a formation form, the safety of the whole formation needs to be considered for the planned path, and the conventional two-dimensional path planning algorithm generally takes the shortest path as a criterion, and the planned path is close to an obstacle, so that the safety of the whole formation cannot be ensured.
Disclosure of Invention
In view of the above analysis, the present invention aims to disclose a ground robot path planning method based on air-ground coordination, which is used to solve the problems that a field high-precision environment map is difficult to obtain, it takes time to construct a map according to a ground robot, and a common path planning algorithm cannot meet the requirements of ground multi-robot path security and performability.
The invention discloses a ground robot path planning method based on air-ground coordination, which comprises the following steps:
s101, constructing a complete space three-dimensional model of a target area based on space-based image acquisition data of the target area;
s102, constructing a moving path area of the ground mobile robot based on the space three-dimensional model, and rasterizing the moving path area to obtain a binary grid map of a path which can be accessed by the robot;
and S103, establishing a mixed map combined with the Voronoi map based on the binary grid map, carrying out mixed A-algorithm global path planning on the mixed map, and carrying out path optimization based on the multi-guide-point traction effect to obtain a reliable moving path of the ground robot.
Further, the step S103 includes:
s401, establishing a Voronoi map by using the binary grid map, and combining the Voronoi map and the binary grid map to construct a mixed map;
step S402, searching out a coarse path which meets the continuity of the path and is based on the ground robot kinematics and the minimum safe steering constraint by adopting mixed A-star search in a mixed map;
and S403, optimizing the coarse path of the hybrid search based on a track smoother of a gradient descent optimization algorithm according to the coarse path to construct a continuous smooth path suitable for safe execution of multiple ground robots.
Further, in step S402, the hybrid a search specifically includes:
step S501, inputting a mixed map;
step S502, initializing searching;
the initialization comprises the steps of establishing an Open _ Set list and a Closed _ Set list, wherein the two lists are empty lists when being initialized, the Open _ Set list stores nodes to be expanded, and the Closed _ Set list stores expanded nodes or nodes with obstacles;
step S503, judging that the Open _ Set list is a non-empty list, selecting a node with the minimum total cost from the Open _ Set list as an expansion node for expansion to obtain a plurality of expansion sub-nodes with continuous states with the expansion node, and removing the expansion node from the Open _ Set list to a Closed _ Set list;
step S504, judge whether the expanded node is the goal point; if yes, ending the search; if not, continuing to expand the child nodes of the node;
step S505, judging whether the expansion sub-node collides with an obstacle or not, if so, discarding the sub-node; if not, jumping to the next step;
step S506, comparing the actual cost of the current expansion child node and the actual cost of the original node, when the actual cost of the current expansion child node is lower than the actual cost of the original node, associating the continuous state of the current expansion child node with the grid unit, setting the grid where the current expansion child node is located as the father grid of the grid where the child node is located, putting the expansion child node into an Open _ Set list, and updating the total cost of the expansion child node;
step S507, looping steps S503 to S506 in the above process until the target node is searched, and placing the target node in the Open _ Set list;
step S508, starting to backtrack all traversed father nodes from the target node in the Open _ Set list until the starting node, obtaining a searched path, and ending the algorithm after the path is searched; if the Open _ Set list is empty, it indicates that no feasible path is planned, and the algorithm terminates.
Further, the total cost of each node in the searching process comprises an actual cost and a heuristic estimation cost;
the actual cost is the cost from the starting node to the current node, and comprises the parent node cost of the current node and the cost from the parent node to the current node, wherein the cost comprises the actual path length from the parent node to the current node, the steering cost from the parent node to the current node in the binary raster map and the cost from the parent node to the current node in the Voronoi graph;
the heuristic estimation cost is the heuristic estimation cost from the current node to the target node and is determined by the larger value of the heuristic function based on the incomplete constraint and the heuristic function based on the complete constraint.
Further, in the process of optimizing the path based on the multi-guide point traction effect, a path smoother based on the multi-guide point effect is adopted to smooth the rough path to obtain a path which is suitable for safe execution and continuous smoothing of the ground robot.
Further, the gradient descent smoother establishes a minimum cost function based on the barrier term, the curvature term, the smoothing term, the voronoi field term and the multi-guiding point function term, and solves the optimal path by using a gradient descent method.
Further, when for each search node x i With | x i -g i |＜ρ max Cost function P of said multiple bootstrap function terms gui Comprises the following steps:
in the formula, x i Two-dimensional plane coordinates of a vertex on the path; g i Is a distance node x i The location of the nearest guidance point; rho max A threshold value of a maximum distance for which the bootstrap point affects the cost function; guide point weight w gui Influencing a road for a guidance pointThe coefficient of the diameter change.
Further, the step S101 includes:
step S201, dividing a target area into a plurality of target sub-areas, acquiring sequence images of the target sub-areas by using at least one unmanned aerial vehicle, and transmitting image scaling maps to a ground server through a map transmission device;
step S202, performing real-time three-dimensional reconstruction in a ground server to recover dense three-dimensional point clouds on the surface of a target sub-region;
and S203, integrally fusing the single unmanned aerial vehicle composition by using a multi-map splicing fusion algorithm based on three-dimensional point cloud data to construct a complete space three-dimensional model of the target area.
Further, in step S102, a deep learning method for predicting a remote sensing image based on a convolutional neural network is adopted to extract a road surface region in the spatial three-dimensional model scene, construct a reliable movement path region of the ground mobile robot, and perform rasterization interpolation on the reliable movement path region to obtain a binary grid map of an accessible path.
Further, the step S102 specifically includes:
s301, making a point cloud data training set;
marking the characteristics including buildings, vegetations, ground, vehicles, pedestrians, lakes, wall surfaces and road beds according to the three-dimensional model of the real scene to construct a data training set;
step S302, establishing a deep semantic segmentation network model for extracting a reliable moving path region;
step S303, training and testing the deep semantic segmentation network model by utilizing the data training set;
step S304, extracting a complete space three-dimensional model of the target area by using the trained deep semantic segmentation network model to obtain a reliable movement path area of the robot in the target area;
and S305, according to the reliable movement path area of the robot, performing rasterization interpolation on the reliable movement path area to obtain a binary grid map of the accessible path.
The invention can realize the following beneficial effects:
the ground robot path planning method based on the air-ground coordination enables the path planned by the robot to be drawn to the vicinity of a plurality of set guide points, and a safe and reliable path is successfully planned in a multi-obstacle environment.
The invention can be suitable for planning the whole path of multi-robot formation, the planned path is close to the middle position of a road, the requirement of the whole width of formation can be met, and the safety of formation movement is ensured.
Because the three-dimensional path planning algorithm is based, the positions and the directions of all path points can be obtained, the path or the trajectory tracking controller is more convenient to design, and the motion control of the robot is more met; the intelligent automobile steering device is not only suitable for intelligent automobiles, but also suitable for other robots with kinematics and steering constraint, avoids unnecessary steering movement, and can reach the specified position in an accurate direction.
The method is suitable for the field of task emergency, such as fire rescue, global information is provided for the ground robot by using the rapid maneuvering capability of the unmanned aerial vehicle, a high-precision on-site three-dimensional space model is rapidly constructed, and then reliable moving path areas including areas such as roads and barriers are extracted, and a reference map is provided for the ground robot to perform path planning.
Drawings
The drawings are only for purposes of illustrating particular embodiments and are not to be construed as limiting the invention, wherein like reference numerals are used to designate like parts throughout.
Fig. 1 is a flowchart of a ground robot path planning method based on air-ground coordination in an embodiment of the present invention;
FIG. 2 is a flowchart of a method for constructing a complete spatial three-dimensional model of a target region according to an embodiment of the present invention;
FIG. 3 is a flowchart of a training and testing process of a deep semantic segmentation network model according to an embodiment of the present invention;
fig. 4 is a flowchart of a method for implementing reliable moving path planning by the improved hybrid a-x algorithm in the embodiment of the present invention;
fig. 5 is a basic flow chart of the hybrid a-algorithm in the embodiment of the present invention;
fig. 6 is a diagram of 3 forward search node expansion modes of the hybrid a-star algorithm in the embodiment of the present invention;
fig. 7 is a diagram of an example of a hybrid a-algorithm search in an embodiment of the present invention.
Detailed Description
The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form a part hereof, and which together with the embodiments of the invention serve to explain the principles of the invention.
One embodiment of the present invention discloses a ground robot path planning method based on air-ground coordination, as shown in fig. 1, including:
s101, constructing a complete space three-dimensional model of a target area based on space-based image acquisition data of the target area;
s102, constructing a moving path area of the ground mobile robot based on the space three-dimensional model, and rasterizing the moving path area to obtain a binary grid map of a path which can be accessed by the robot;
and S103, establishing a mixed map combined with the Voronoi map based on the binary grid map, carrying out mixed A-algorithm global path planning on the mixed map, and carrying out path optimization based on the multi-guide-point traction effect to obtain a reliable moving path of the ground robot.
As shown in fig. 2, step S101 specifically includes:
step S201, dividing a target area into a plurality of target sub-areas, acquiring sequence images of the target sub-areas by using at least one unmanned aerial vehicle, and transmitting image scaling maps to a ground server through a map transmission device;
because the sensing capability and the maneuvering capability of the ground robot are usually limited, when the ground robot is in a large environment scene such as a fire rescue in a large field environment, if a camera or a laser radar is carried by one or more ground robots to carry out global mapping on a field area, a lot of time is consumed, and the method is not very suitable for the emergency fire rescue process. And the unmanned aerial vehicle has a wide aerial view field and quick maneuvering capability.
The method comprises the steps of determining the flight track of at least one unmanned aerial vehicle for shooting target area images according to the task site range, the real-time requirement of drawing construction and the number of available unmanned aerial vehicles, selecting the flight track as the starting position of the ground robot to the task target end position in order to construct a reliable moving area map for the ground robot to walk as soon as possible, efficiently and accurately acquiring sequence images of target sub-areas during flight, and transmitting an image scaling map to a ground server through an image transmission device.
S202, performing real-time three-dimensional reconstruction in a ground server to recover dense three-dimensional point clouds on the surface of a target sub-region;
in this embodiment, a mature application real-time three-dimensional reconstruction technology can be adopted, so that the real-time three-dimensional reconstruction of the unmanned aerial vehicle on the task site becomes practical.
Step S203, performing integral fusion on the single unmanned aerial vehicle composition by using a multi-map splicing fusion algorithm based on three-dimensional point cloud data, thereby efficiently constructing a complete space three-dimensional model of a target area;
extracting edge pixel points of a point cloud map to be spliced according to a three-dimensional point cloud map of a sub-target area acquired by each unmanned aerial vehicle, establishing a pixel point transformation mathematical model, converting a map splicing problem into data splicing and registration of the three-dimensional point cloud, and obtaining a translation rotation matrix of the pixel points; and according to the pixel point transformation result, performing corresponding rotation, translation and fusion on the map to be spliced to obtain an accurate splicing result of the complete space three-dimensional model of the target area.
Specifically, in step S102, a deep learning method for predicting a remote sensing image based on a convolutional neural network is adopted to extract a road surface region in the spatial three-dimensional model scene, construct a reliable movement path region of the ground mobile robot, and perform rasterization interpolation on the reliable movement path region to obtain a binary grid map capable of reaching a path;
in step S102, the binary grid map of the robot reachable path includes:
s301, making a point cloud data training set;
according to the three-dimensional model of the real scene, marking the characteristics including buildings, vegetation, the ground, vehicles, pedestrians, lakes, wall surfaces and road beds to construct a data training set;
step S302, establishing a deep semantic segmentation network model for extracting a reliable moving path region;
in the present embodiment, an image semantic segmentation model of an encoder-decoder model of a classical structure is adopted as a deep semantic segmentation network model for performing reliable movement path region extraction.
Step S303, training and testing the deep semantic segmentation network model by utilizing the data training set;
and training the established deep semantic segmentation network model by using the training data set, and inputting the test data of the established complete target region three-dimensional model into the trained deep semantic segmentation network model to obtain the extraction result of the reliable moving path region of the remote sensing image.
Step S304, extracting a complete space three-dimensional model of a target area by using the trained deep semantic segmentation network model to obtain a reliable movement path area of the robot in the target area;
and S305, performing rasterization interpolation according to the reliable movement path area of the robot to obtain a binary raster map of the accessible path.
The rasterization interpolation can be performed by adopting the existing map rasterization method.
FIG. 3 is a flowchart of a training and testing process of a deep semantic segmentation network model;
as shown in fig. 4, in step S103, the method specifically includes:
step S401, establishing a Voronoi map by using a binary grid map, and combining the Voronoi map and the binary grid map to construct a mixed map;
one disadvantage of conventional path planning algorithms is that the path it plans is too close to the obstacle, i.e. it selects the shortest path without collision. In order to trade off the contradiction between shortest path and far away from an obstacle, one common approach is to use a potential field to penalize the robot to approach the obstacle, however, the traditional potential field has several serious drawbacks.
First, conventional potential fields tend to create high potential regions in narrow channels, which makes the path traversing these channels computationally expensive.
Secondly, since the applied potential field around an obstacle is usually defined as a function of the distance from the obstacle, this means that calculating a given potential field value requires calculating the potential energy encompassing all obstacles within the effective radius, which can be computationally expensive.
To solve these problems, the present embodiment introduces Voronoi fields, readjusting the potential field distribution according to the geometry of the binary rasterized map.
Specifically, in a binary grid map, a certain distance threshold value from an obstacle is selected by using a Voronoi potential field function, and a two-dimensional Voronoi map of a field space is constructed;
the Voronoi field is used to define a trade-off relationship between path length and proximity to obstacles. The Voronoi potential field function is defined as follows:
in the formula (d) O And d V Distance to the nearest obstacle and distance to the edge of the nearest generalized Tassen polygon (GVD), respectively, a ∈ [0, ∞),
the falling rate of the potential energy value and the control range are respectively controlled for the parameters required to be adjusted. When the temperature is higher than the set temperature
The expression (1) holds; otherwise, ρ V (x,y)＝0。
The potential energy field has the following properties:
(i) When in use
The potential energy is 0.
(ii) Potential energy value ρ V (x,y)∈[0,1]And the distribution is continuous.
(iii) The potential energy value reaches a maximum when a point (x, y) on the map is on or inside an obstacle.
(iv) When a point (x, y) on the map is on the side of the generalized voronoi diagram, its potential energy value reaches a minimum value.
A key advantage of Voronoi fields compared to conventional potential fields is that the value of the field can be scaled in proportion to the potential field of navigation. Thus, even a narrow U-shaped opening can navigate, which is not always the case for a conventional potential field. In addition, the potential field at the middle position of the channel is zero, so that a reliable and safe path can be provided for multi-robot formation driving.
Preferably, the Voronoi diagram is generated based on ROS (robotic operating system) using a constructed Voronoi potential field function.
And superposing and combining the Voronoi map and the binary grid map to construct a mixed map.
And S402, searching out a coarse path which meets the continuity of the path based on the ground robot kinematics and the minimum safe steering constraint by adopting mixed A-star search in the mixed map.
The A-algorithm in the heuristic search algorithm is relatively universal, but the hybrid A-algorithm is adopted as the search technology of the global path in a variant form of the A-algorithm in consideration of the defect that the vehicle kinematic constraint of the fire-fighting robot has discreteness with the traditional A-algorithm.
Compared with a traditional A-algorithm two-dimensional search space comprising position information in the x direction and the y direction, the spatial dimension of the hybrid A-algorithm search is four-dimensional, wherein the information representing the current orientation of the robot is added, and the mode that the fourth dimension represents the forward and backward movement of the robot, the forward movement or the backward movement is added.
The traditional a-algorithm has the characteristic of discrete state, so that the planned path cannot be directly executed by the robot. To overcome this problem, blend a is assigned to the corresponding discrete grid with the states of the continuous robot, so that the connected continuous coordinates will be performed by the actual robot.
The hybrid A-algorithm is a search algorithm based on a heuristic function, path search is carried out from a starting point to a target point in an expanded node mode, and the total cost of each node in the search process is measured by utilizing an evaluation function. The evaluation function is f(s) = g(s) + h(s), wherein g(s) is expressed as the actual cost from the starting node to the current node, and the following recursion relation is satisfied:
g(s i )＝g(s i-1 )+cost(s k-1 ,s k )
＝g(s i-1 )+(1+a·turncost(s k-1 ,s k )+b·mapcost(s k-1 ,s k ))·dist(s k-1 ,s k )
in the formula, s i Is the current node (child node), s i-1 Is the previous node (parent node); g(s) i ) G(s) as the actual cost of the current node i-1 ) Actual cost of parent node of current node; cost(s) k-1 ,s k ) Is the cost from the parent node to the current node; a and b are each weight factor; dist(s) k-1 ,s k ) Is the path length from the parent node to the current node; turncost(s) k-1 ,s k ) The value is 0 when going straight and 1 when turning, which is the cost of turning from the parent node to the current node. The action is to punish the turning of the planned path, so that the robot can keep going straight as much as possible; mappost(s) k-1 ,s k ) For the cost of the Voronoi diagram from the parent node to the current node, the value is infinite within the barrier safety distance, otherwise it is 0. The function is to ensure that the planned path avoids obstacles and the running safety of the robot is ensured.
The second term h(s) represents the heuristic estimation cost from the current node to the target node and is determined by the larger value of the heuristic function based on the incomplete constraint and the heuristic function based on the complete constraint, and the expression is h(s) i )＝max h 1 (s i ),h 2 (s i ) Wherein h is 1 In order to obtain a heuristic value under the conditions of ignoring environmental obstacles and considering the non-integrity constraint of the robot, the path length obtained by Reeds-Shepp or Dubins is generally selected; h is a total of 2 In order to obtain a heuristic value under the condition of ignoring the non-integrity constraint of the robot and considering the environment constraint, the path length obtained by the A-algorithm search is generally selected.
Specifically, the basic flow of the hybrid a-algorithm is shown in fig. 5: the method comprises the following steps:
step S501, inputting a mixed map;
step S502, initialization is carried out;
the initialization comprises the steps of establishing an Open _ Set list and a Closed _ Set list, wherein the two lists are empty lists when the two lists are initialized, the Open _ Set list stores nodes to be expanded, and the Closed _ Set list stores expanded nodes or nodes with obstacles. Obtaining a starting point s 0 (x 0 ,y 0 ,θ 0 ) And target point s g (x g ,y g ,θ g ) And a kinematic model of the differential mobile robot. Putting the starting point into an Open _ Set list;
step S503, judging that the Open _ Set list is a non-empty list, selecting a node with the minimum total cost from the Open _ Set list as an expansion node for expansion to obtain a plurality of expansion sub-nodes with continuous states with the expansion node, and removing the expansion node from the Open _ Set list to the Closed _ Set list;
when the retrieved Open _ Set list is an empty list, the path search fails, and the map has no path to the target position. If the list is not an empty list, selecting the node s with the minimum total cost from the Open _ Set list i (x i ,y i ,θ i ) As an extension node, 3 child nodes continuous with the node state are extended, and the node s is extended i Moving the Open _ Set list into a Closed _ Set list; x is said i ,y i Is a coordinate value of θ i Is the direction angle.
Preferably, the expansion directions are three directions, namely, left rotation, right rotation and straight movement of the node in the forward movement direction, and three extended continuous sub-nodes are shown in fig. 6. For the linear node, the robot moves along the linear motion direction; for the turning node, the robot moves along the arc motion direction, and the arc length slightly larger than the minimum turning radius of the robot can be adopted during expansion.
Step S504, judging the extension node S i Whether it is the target point s g (ii) a If yes, ending the search; if not, continuing to expand the node s i S of i+1 ；
Step S505 of judging the expansion child node S i+1 If not, the child node is abandoned; if not, jumping to the next step;
step S506, the current expansion child node S appearing in the same grid i+1 With the original node s open Is compared with the actual cost g(s) of the current extended child node i+1 ) Lower than the actual cost g(s) of the original node open ) Then, the continuous state of the current expansion child node is associated with the grid cell, and the current expansion child node s is associated with the grid cell i+1 Setting the grid to be the father grid of the grid where the child node is positioned, placing the extended child node into an Open _ Set list, and updating the total cost f(s) of the extended child node i+1 ) (ii) a Otherwise, the currently searched child node is discarded.
Step S507, looping steps S503 to S506 in the above process until the target node is searched, and placing the target node in the Open _ Set list;
step S508, starting to backtrack all traversed father nodes from the target node in the Open _ Set list until the starting node, obtaining a searched path, and ending the algorithm after the path is searched; if the Open _ Set list is empty, it indicates that no feasible path is planned, and the algorithm terminates.
Fig. 7 shows a search example of the hybrid a-star algorithm by taking the expansion of 3 forward nodes as an example.
And S403, optimizing the coarse path of the hybrid search based on a track smoother of a gradient descent optimization algorithm according to the coarse path to construct a continuous smooth path suitable for safe execution of multiple ground robots.
Although the hybrid a-star algorithm can output a continuous path for the robot to execute, the robot executing process usually includes many unnecessary turning actions, and in addition, in a scene that the robot carries out fire extinguishing tasks by using the towed hose, the problem of winding of obstacles in the environment on the robot hose during towing needs to be solved.
Therefore, in this embodiment, a path smoother based on multi-boot point is disclosed, and a safer, more reliable and better path is obtained through processing.
The gradient descent smoother establishes a minimum cost function based on an obstacle term, a curvature term, a smoothing term, a voronoi field term and a multi-guide-point function term, and solves an optimal path by using a gradient descent method.
Specifically, the minimized cost function P of the gradient descent smoother is:
P＝P obs +P cur +P smo +P vor +P gui (2)
(1) Item of obstacle
When for each search node x i With | x i -o i |≤d max Defining a cost function P obs Comprises the following steps:
in the formula, N is the total number of nodes; x is the number of i Two-dimensional plane coordinates of a vertex on the path; o i Is a distance node x i The location of the nearest obstacle; d max A threshold value of maximum distance for which an obstacle affects the cost function; sigma obs A secondary penalty function is adopted, so that the penalty value is larger when the node is closer to the barrier; weight of obstacle w obs Is the coefficient that the obstacle affects the path change. This item enables the robot to effectively avoid collisions with obstacles.
(2) Term of curvature
To ensure that the path is executable, the curvature term sets an upper limit on the instantaneous curvature of each vertex, i.e., assume
Defining a cost functionNumber P cur Comprises the following steps:
in the formula, the vertex x i The displacement vector of (b) is defined as Δ x i ＝x i -x i-1 The change value of the vertex tangential angle is expressed as
Maximum allowed curvature is represented by k max Represents; sigma cur Is a secondary penalty function; curvature weight w cur Controlling the effect on the path change. This term limits the instantaneous curvature of the trajectory at each node and enforces incomplete constraint of the vehicle.
(3) Smoothing term
The smoothing term evaluates the displacement vectors between the vertices. The term can smooth nodes with uneven intervals and large direction change amplitude. Defining a cost function P smo Comprises the following steps:
in the formula, w smo The smoothing weight is expressed and determines the influence of the smoothing term on the path change.
(4) Voronoi potential field function term
For
Defining a cost function P vor Is composed of
In the formula (I), the compound is shown in the specification, dO representing the distance of the node to the nearest obstacle; d edg Representing the distance from the node to the nearest edge of the generalized Thiessen polygon;
representing the maximum distance of an obstacle affecting the Voronoi potential field function; the decay rate of the alpha control field; w is a vor Is a weight coefficient representing the influence on the path. This term can effectively guide the path away from obstacles in narrow or wide passageways. Closer to the obstacle, P vor The larger the value of (a), the closer to 1; closer to voronoi edge, P vor The closer to 0.
(5) Multiple guide point function terms
When for each search node x i With | x i -g i |＜ρ max Defining a cost function P gui Comprises the following steps:
in the formula, x i Two-dimensional plane coordinates of a vertex on the path; g i Is a distance node x i The location of the nearest guidance point; rho max A threshold value of a maximum distance for which the bootstrap point affects the cost function; guide point weight w gui Coefficients that influence path changes for the guide points. Distance rho when path search node is closer to guide point max Boundary of (P) gui The greater the value of (A); the closer to the guide point, P gui The smaller the value of (c). This term can lead quickly to the vicinity of the guidance point by the path node.
After the cost function P is determined, the optimal path is solved by using a gradient descent method, in the actual application of the gradient descent algorithm, the absolute value of the gradient is usually selected as the standard for stopping the algorithm, and the continuity of the motion of the robot is ensured by limiting the maximum iteration number.
In summary, according to the ground robot path planning method based on the air-ground coordination in the embodiment of the invention, a more efficient and accurate space three-dimensional model in a large-scale environment is established through the air-ground image acquisition data; a binary grid map is extracted through a deep semantic segmentation network model, and an improved hybrid A-x algorithm is adopted to plan a path based on the binary grid map, so that the path planned by the robot can be drawn to the vicinity of a plurality of set guide points, and a safe and reliable path is successfully planned in a multi-obstacle environment.
The embodiment of the invention can be suitable for planning the whole path of multi-robot formation, the planned path is close to the middle position of a road, the requirement on the whole width of formation can be met, and the safety of formation movement is ensured.
Because the position and the direction of all path points can be obtained based on a three-dimensional path planning algorithm, the method is more convenient when a path or a track tracking controller is designed, and is more in line with the motion control of a robot; the intelligent steering mechanism is not only suitable for intelligent automobiles, but also suitable for other robots with motion and steering constraints, avoids unnecessary steering motion, and can reach the designated position in an accurate direction.
The method is suitable for the field of task emergency, such as fire rescue, global information is provided for the ground robot by using the rapid maneuvering capability of the unmanned aerial vehicle, a high-precision on-site three-dimensional space model is rapidly constructed, and then reliable moving path areas including areas such as roads and barriers are extracted, and a reference map is provided for the ground robot to perform path planning.
While the invention has been described with reference to specific preferred embodiments, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the following claims.
Claims (10)
1. A ground robot path planning method based on air-ground coordination is characterized by comprising the following steps:
s101, constructing a complete space three-dimensional model of a target area based on space-based image acquisition data of the target area;
s102, constructing a moving path area of the ground mobile robot based on the space three-dimensional model, and rasterizing the moving path area to obtain a binary grid map of a path which can be accessed by the robot;
and S103, establishing a mixed map combined with the Voronoi map based on the binary grid map, carrying out mixed A-algorithm global path planning on the mixed map, and carrying out path optimization based on the multi-guide-point traction effect to obtain a reliable moving path of the ground robot.
2. The ground robot path planning method according to claim 1, wherein the step S103 includes:
step S401, establishing a Voronoi map by using a binary grid map, and combining the Voronoi map and the binary grid map to construct a mixed map;
step S402, searching out a coarse path which meets the continuity of the path based on the ground robot kinematics and the minimum safe steering constraint by adopting a mixed A-search in a mixed map;
and S403, optimizing the coarse path of the hybrid search based on a track smoother of a gradient descent optimization algorithm according to the coarse path to construct a continuous smooth path suitable for safe execution of multiple ground robots.
3. The ground robot path planning method according to claim 2, wherein in the step S402, the hybrid a search specifically includes:
step S501, inputting a mixed map;
step S502, initializing searching;
the initialization comprises the steps of establishing an Open _ Set list and a Closed _ Set list, wherein the two lists are empty lists during initialization, the Open _ Set list stores nodes to be expanded, and the Closed _ Set list stores expanded nodes or nodes with obstacles;
step S503, judging that the Open _ Set list is a non-empty list, selecting a node with the minimum total cost from the Open _ Set list as an expansion node for expansion to obtain a plurality of expansion sub-nodes with continuous states with the expansion node, and removing the expansion node from the Open _ Set list to a Closed _ Set list;
step S504, judge whether the expanded node is the goal point; if yes, ending the search; if not, continuing to expand the child nodes of the node;
step S505, judging whether the expansion sub-node collides with an obstacle or not, if so, discarding the sub-node; if not, jumping to the next step;
step S506, comparing the actual cost of the current extended child node and the original node in the same grid, and when the actual cost of the current extended child node is lower than the actual cost of the original node, associating the continuous state of the current extended child node with the grid unit, setting the grid in which the current extended child node is positioned as the father grid of the grid in which the child node is positioned, putting the extended child node into an Open _ Set list, and updating the total cost of the extended child node;
step S507, looping steps S503 to S506 in the above process until the target node is searched, and placing the target node in an Open _ Set list;
step S508, starting to backtrack all traversed father nodes from the target node in the Open _ Set list until the starting node, obtaining a searched path, and ending the algorithm after the path is searched; if the Open _ Set list is empty, it indicates that no feasible path is planned, and the algorithm terminates.
4. The ground robot path planning method according to claim 2,
the total cost of each node in the searching process comprises actual cost and heuristic estimation cost;
the actual cost is the cost from the starting node to the current node, and comprises the parent node cost of the current node and the cost from the parent node to the current node, wherein the cost comprises the actual path length from the parent node to the current node, the steering cost from the parent node to the current node in the binary raster map and the cost from the parent node to the current node in the Voronoi graph;
the heuristic estimation cost is the heuristic estimation cost from the current node to the target node and is determined by the larger value of the heuristic function based on the incomplete constraint and the heuristic function based on the complete constraint.
5. The ground robot path planning method of claim 2,
and in the process of optimizing the path based on the multi-guide-point traction effect, a path smoother based on the multi-guide-point effect is adopted to smooth the rough path to obtain a path which is suitable for safe execution and continuous smoothing of the ground robot.
6. The ground robot path planning method of claim 2,
the gradient descent smoother establishes a minimum cost function based on an obstacle term, a curvature term, a smoothing term, a voronoi field term and a multi-guide-point function term, and solves an optimal path by using a gradient descent method.
7. The ground robot path planning method according to claim 2,
when for each search node x i With | x i -g i |＜ρ max Cost function P of said multiple bootstrap function terms gui Comprises the following steps:
in the formula, x i Two-dimensional plane coordinates of a vertex on the path; g i Is a distance node x i The location of the nearest guidance point; ρ is a unit of a gradient max A threshold value of a maximum distance for which the bootstrap point affects the cost function; guide point weight w gui Coefficients that influence the path change for the guidance point.
8. The ground robot path planning method of claim 2,
the step S101 includes:
step S201, dividing a target area into a plurality of target sub-areas, acquiring sequence images of the target sub-areas by using at least one unmanned aerial vehicle, and transmitting an image scaling map to a ground server through a map transmission device;
s202, performing real-time three-dimensional reconstruction in a ground server to recover dense three-dimensional point clouds on the surface of a target sub-region;
and S203, integrally fusing the single unmanned aerial vehicle composition by using a multi-map splicing fusion algorithm based on three-dimensional point cloud data to construct a complete space three-dimensional model of the target area.
9. The ground robot path planning method of claim 2,
in step S102, a deep learning method for predicting a remote sensing image based on a convolutional neural network is adopted to extract a road surface region in the spatial three-dimensional model scene, construct a reliable movement path region of the ground mobile robot, and perform rasterization interpolation on the reliable movement path region to obtain a binary raster map of an accessible path.
10. The ground robot path planning method according to claim 9,
the step S102 specifically includes:
s301, making a point cloud data training set;
marking the characteristics including buildings, vegetations, ground, vehicles, pedestrians, lakes, wall surfaces and road beds according to the three-dimensional model of the real scene to construct a data training set;
step S302, establishing a deep semantic segmentation network model for extracting a reliable moving path region;
step S303, training and testing the deep semantic segmentation network model by utilizing the data training set;
step S304, extracting a complete space three-dimensional model of a target area by using the trained deep semantic segmentation network model to obtain a reliable movement path area of the robot in the target area;
and S305, according to the reliable movement path area of the robot, performing rasterization interpolation on the reliable movement path area to obtain a binary grid map of the accessible path.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211108617.6A CN115373399A (en) | 2022-09-13 | 2022-09-13 | Ground robot path planning method based on air-ground cooperation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211108617.6A CN115373399A (en) | 2022-09-13 | 2022-09-13 | Ground robot path planning method based on air-ground cooperation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115373399A true CN115373399A (en) | 2022-11-22 |
Family
ID=84072481
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211108617.6A Pending CN115373399A (en) | 2022-09-13 | 2022-09-13 | Ground robot path planning method based on air-ground cooperation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115373399A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115562255A (en) * | 2022-09-13 | 2023-01-03 | 中国安全生产科学研究院 | Multi-intelligent fire-fighting robot fire hose anti-winding method based on air-ground cooperation |
| CN115930969A (en) * | 2023-01-09 | 2023-04-07 | 季华实验室 | Path planning method and device for mobile robot, electronic equipment and storage medium |
| CN116311095A (en) * | 2023-03-16 | 2023-06-23 | 广州市衡正工程质量检测有限公司 | Pavement detection method based on region division, computer equipment and storage medium |
| CN116433865A (en) * | 2023-06-15 | 2023-07-14 | 深圳大学 | Space-ground collaborative acquisition path planning method based on scene reconstructability analysis |
| CN116772880A (en) * | 2023-06-07 | 2023-09-19 | 中国人民解放军陆军装甲兵学院 | Unmanned aerial vehicle path planning method based on unmanned aerial vehicle vision |
| CN116909268A (en) * | 2023-06-30 | 2023-10-20 | 广东省机场管理集团有限公司工程建设指挥部 | 5G-based path planning method, device, equipment and medium for walking robot |
| CN117685994A (en) * | 2024-02-04 | 2024-03-12 | 北京航空航天大学 | Unmanned vehicle path planning method for air-ground coordination |
| CN117930855A (en) * | 2024-03-25 | 2024-04-26 | 山东工程职业技术大学 | Robot inspection method and system for construction site |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106647769A (en) * | 2017-01-19 | 2017-05-10 | 厦门大学 | AGV path tracking and obstacle avoiding coordination method based on A* extraction guide point |
| CN108898605A (en) * | 2018-07-25 | 2018-11-27 | 电子科技大学 | A kind of grating map dividing method based on figure |
| CN111323037A (en) * | 2020-02-28 | 2020-06-23 | 武汉科技大学 | Voronoi path planning algorithm for novel framework extraction of mobile robot |
| CN111998859A (en) * | 2020-09-16 | 2020-11-27 | 杭州海康消防科技有限公司 | Path planning method and device, electronic equipment and path indicating system |
| CN113504793A (en) * | 2021-07-12 | 2021-10-15 | 哈尔滨工程大学 | Unmanned aerial vehicle cluster path planning method based on Voronoi diagram |
| CN113759971A (en) * | 2021-08-30 | 2021-12-07 | 中国人民解放军国防科技大学 | Path planning method for unmanned aerial vehicle cooperative reconnaissance |
| CN114237232A (en) * | 2021-11-30 | 2022-03-25 | 武汉地大信息工程股份有限公司 | Path planning method and device |
| CN115032999A (en) * | 2022-06-24 | 2022-09-09 | 中国安全生产科学研究院 | Dynamic optimization formation transformation method for multi-robot formation |
-
2022
- 2022-09-13 CN CN202211108617.6A patent/CN115373399A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106647769A (en) * | 2017-01-19 | 2017-05-10 | 厦门大学 | AGV path tracking and obstacle avoiding coordination method based on A* extraction guide point |
| CN108898605A (en) * | 2018-07-25 | 2018-11-27 | 电子科技大学 | A kind of grating map dividing method based on figure |
| CN111323037A (en) * | 2020-02-28 | 2020-06-23 | 武汉科技大学 | Voronoi path planning algorithm for novel framework extraction of mobile robot |
| CN111998859A (en) * | 2020-09-16 | 2020-11-27 | 杭州海康消防科技有限公司 | Path planning method and device, electronic equipment and path indicating system |
| CN113504793A (en) * | 2021-07-12 | 2021-10-15 | 哈尔滨工程大学 | Unmanned aerial vehicle cluster path planning method based on Voronoi diagram |
| CN113759971A (en) * | 2021-08-30 | 2021-12-07 | 中国人民解放军国防科技大学 | Path planning method for unmanned aerial vehicle cooperative reconnaissance |
| CN114237232A (en) * | 2021-11-30 | 2022-03-25 | 武汉地大信息工程股份有限公司 | Path planning method and device |
| CN115032999A (en) * | 2022-06-24 | 2022-09-09 | 中国安全生产科学研究院 | Dynamic optimization formation transformation method for multi-robot formation |
Non-Patent Citations (3)
| Title |
|---|
| 张彪，等: "使用三维栅格地图的移动机器人路径规划", 西安交通大学学报, vol. 47, no. 10, 31 October 2013 (2013-10-31), pages 57 - 61 * |
| 车洪磊，等: "基于改进蚁群算法的应急监测移动机器人全局路径规划", 中国安全生产科学技术, vol. 12, 30 September 2016 (2016-09-30), pages 278 - 183 * |
| 陈龚，等: "无人机协同路径规划三维实时可视化仿真平台", ***仿真学报, vol. 28, no. 9, 30 September 2016 (2016-09-30), pages 1996 - 2001 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115562255A (en) * | 2022-09-13 | 2023-01-03 | 中国安全生产科学研究院 | Multi-intelligent fire-fighting robot fire hose anti-winding method based on air-ground cooperation |
| CN115930969A (en) * | 2023-01-09 | 2023-04-07 | 季华实验室 | Path planning method and device for mobile robot, electronic equipment and storage medium |
| CN116311095A (en) * | 2023-03-16 | 2023-06-23 | 广州市衡正工程质量检测有限公司 | Pavement detection method based on region division, computer equipment and storage medium |
| CN116311095B (en) * | 2023-03-16 | 2024-01-02 | 广州市衡正工程质量检测有限公司 | Pavement detection method based on region division, computer equipment and storage medium |
| CN116772880B (en) * | 2023-06-07 | 2024-01-23 | 中国人民解放军陆军装甲兵学院 | Unmanned aerial vehicle path planning method based on unmanned aerial vehicle vision |
| CN116772880A (en) * | 2023-06-07 | 2023-09-19 | 中国人民解放军陆军装甲兵学院 | Unmanned aerial vehicle path planning method based on unmanned aerial vehicle vision |
| CN116433865B (en) * | 2023-06-15 | 2023-10-20 | 深圳大学 | Space-ground collaborative acquisition path planning method based on scene reconstructability analysis |
| CN116433865A (en) * | 2023-06-15 | 2023-07-14 | 深圳大学 | Space-ground collaborative acquisition path planning method based on scene reconstructability analysis |
| CN116909268A (en) * | 2023-06-30 | 2023-10-20 | 广东省机场管理集团有限公司工程建设指挥部 | 5G-based path planning method, device, equipment and medium for walking robot |
| CN116909268B (en) * | 2023-06-30 | 2024-05-28 | 广东省机场管理集团有限公司工程建设指挥部 | 5G-based path planning method, device, equipment and medium for walking robot |
| CN117685994A (en) * | 2024-02-04 | 2024-03-12 | 北京航空航天大学 | Unmanned vehicle path planning method for air-ground coordination |
| CN117685994B (en) * | 2024-02-04 | 2024-05-17 | 北京航空航天大学 | Unmanned vehicle path planning method for air-ground coordination |
| CN117930855A (en) * | 2024-03-25 | 2024-04-26 | 山东工程职业技术大学 | Robot inspection method and system for construction site |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115373399A (en) | Ground robot path planning method based on air-ground cooperation | |
| Dang et al. | Graph-based path planning for autonomous robotic exploration in subterranean environments | |
| Tai et al. | Virtual-to-real deep reinforcement learning: Continuous control of mobile robots for mapless navigation | |
| Papachristos et al. | The power-tethered UAV-UGV team: A collaborative strategy for navigation in partially-mapped environments | |
| Reinhart et al. | Learning-based path planning for autonomous exploration of subterranean environments | |
| Ström et al. | Predictive exploration considering previously mapped environments | |
| CN109917817B (en) | Multi-underwater robot collaborative path planning method | |
| Jain et al. | Autonomous river exploration | |
| CN112947594B (en) | Unmanned aerial vehicle-oriented track planning method | |
| Zhang et al. | P-CAP: Pre-computed alternative paths to enable aggressive aerial maneuvers in cluttered environments | |
| CN115729266A (en) | Unmanned aerial vehicle obstacle avoidance method for effectively tracking preset air route | |
| Song et al. | A new hybrid method in global dynamic path planning of mobile robot | |
| CN111176281A (en) | Multi-surface unmanned ship coverage type collaborative search method and system based on quadrant method | |
| CN117406771B (en) | Efficient autonomous exploration method, system and equipment based on four-rotor unmanned aerial vehicle | |
| Wu et al. | Real-time three-dimensional smooth path planning for unmanned aerial vehicles in completely unknown cluttered environments | |
| Rocha et al. | Plannie: A benchmark framework for autonomous robots path planning algorithms integrated to simulated and real environments | |
| Short et al. | Abio-inspiredalgorithminimage-based pathplanning and localization using visual features and maps | |
| Politi et al. | Path planning and landing for unmanned aerial vehicles using ai | |
| Collins et al. | Efficient planning for high-speed mav flight in unknown environments using online sparse topological graphs | |
| JP2020149095A (en) | Inverted pendulum robot | |
| Adolf | Multi-query path planning for exploration tasks with an unmanned rotorcraft | |
| Gao et al. | Design and Implementation of an Autonomous Driving Delivery Robot | |
| Wang et al. | Real-time path planning algorithm for autonomous vehicles in unknown environments | |
| Meng et al. | A navigation framework for mobile robots with 3D LiDAR and monocular camera | |
| CN115562255A (en) | Multi-intelligent fire-fighting robot fire hose anti-winding method based on air-ground cooperation |
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 |