Path planning method and device, electronic equipment and computer storage medium

Abstract

The invention provides a path planning method and device, electronic equipment and a computer storage medium, and the method comprises the steps: carrying out the preliminary path planning in a Frenet coordinate system of a reference path, and obtaining an initial path; for each obstacle in the initial path, generating a collision envelope of the vehicle according to the type of the obstacle; if the collision envelope of the vehicle on the initial path is overlapped with the obstacle, taking the current path point as a target path point; then, constructing a drivable tunnel for the target path point to obtain the drivable tunnel of the target path point; and finally, optimizing the initial path according to the drivable tunnels of all the target path points and a preset constraint condition to obtain a final path. Therefore, the purpose of efficiently and stably planning the obstacle avoidance path for different types of obstacles is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of path planning, and particularly relates to a path planning method and device, electronic equipment and computer storage medium. BACKGROUND

[0002] Currently, an automatic driving system needs to avoid obstacles to ensure the safety and efficiency of system operation. However, simply avoiding obstacles in the form of detouring obstacles will affect the feasibility of path planning and the efficiency of system operation. For example, when there is a small obstacle on the road in front, only detouring obstacles to avoid obstacles may have problems such as road conditions not allowing detouring obstacles and detouring obstacles reducing efficiency. SUMMARY

[0003] Therefore, the present application provides a path planning method, device, electronic equipment and computer storage medium, which can efficiently and stably plan an obstacle avoidance path when facing different obstacle types.

[0004] The first aspect of the present application provides a path planning method, comprising:

[0005] performing preliminary path planning in a Frenet coordinate system of a reference path to obtain an initial path;

[0006] for each obstacle in the initial path, generating a collision envelope of a vehicle according to a type of the obstacle;

[0007] determining whether the collision envelope of the vehicle overlaps with the obstacle on the initial path;

[0008] if it is determined that the collision envelope of the vehicle overlaps with the obstacle on the initial path, taking a current path point as a target path point;

[0009] constructing a drivable tunnel for the target path point to obtain a drivable tunnel of the target path point;

[0010] optimizing the initial path according to the drivable tunnels of all the target path points and a preset constraint condition to obtain a final path.

[0011] Optionally, the preliminary path planning in the Frenet coordinate system of the reference path to obtain the initial path comprises:

[0012] equally dividing a longitudinal axis of the Frenet coordinate system of the reference path into N sampling layers; wherein N is a positive integer;

[0013] for each sampling layer, equally sampling M sampling points on the sampling layer from one side boundary to another side boundary along a transverse axis of the Frenet coordinate system of the reference path; wherein M is a positive integer.

[0014] calculating a cost of each sampling point of each sampling layer and recording a parent node of the sampling point;

[0015] starting from the sampling point with the minimum cost in the last sampling layer, tracing back to the starting point along the parent nodes of the sampling points to the previous sampling layer, and obtaining an initial path.

[0016] Optionally, the calculating the cost of each sampling point of each sampling layer comprises:

[0017] determining the cost of the sampling point according to a lateral displacement cost, a lateral offset change cost, a lateral offset change rate cost, and a collision cost of the sampling point.

[0018] Optionally, if there is a cross-lane obstacle in the reference path, the sampling M sampling points on one side boundary to the other side boundary of the lateral axis of the Frenet coordinate system of the reference path for each sampling layer further comprises:

[0019] determining the nearest sampling point of the cross-lane obstacle on the Frenet coordinate system of the reference path according to the coordinates of the cross-lane obstacle in the Frenet coordinate system of the reference path.

[0020] replacing the lateral coordinate of the nearest sampling point with the lateral coordinate of the cross-lane obstacle in the Frenet coordinate system of the reference path.

[0021] Optionally, the determining the nearest sampling point of the cross-lane obstacle on the Frenet coordinate system of the reference path according to the coordinates of the cross-lane obstacle in the Frenet coordinate system of the reference path comprises:

[0022] determining the nearest sampling layer of the cross-lane obstacle according to the longitudinal coordinate of the cross-lane obstacle in the Frenet coordinate system of the reference path.

[0023] calculating the distance between each sampling point in the nearest sampling layer and the cross-lane obstacle.

[0024] taking the sampling point closest to the cross-lane obstacle as the nearest sampling point of the cross-lane obstacle on the Frenet coordinate system of the reference path.

[0025] Optionally, the one side boundary and the other side boundary of the lateral axis of the Frenet coordinate system of the reference path leave a sampling threshold.

[0026] Optionally, the generating, for each of the obstacles in the initial path, a collision envelope of the vehicle according to a type of the obstacle comprises:

[0027] The generating, for each of the obstacles in the initial path, a collision envelope of the vehicle according to a type of the obstacle comprises:

[0028] The generating, for each of the obstacles in the initial path, a collision envelope of the vehicle according to a type of the obstacle comprises:

[0029] Optionally, the constructing, for the target path point, a drivable tunnel to obtain a drivable tunnel of the target path point comprises:

[0030] The generating, for the target path point, a first disc and a second disc according to size information of the vehicle and coordinate information of a center of a rear axle of the vehicle, wherein the first disc and the second disc have the same radius.

[0031] The inflating the obstacle according to the radius of the first disc and the second disc to obtain an inflated obstacle.

[0032] The generating, according to a center of the first disc, the inflated obstacle and a preset step length, a first boundary of the drivable tunnel.

[0033] The generating, according to a center of the second disc, the inflated obstacle and a preset step length, a second boundary of the drivable tunnel.

[0034] The generating, according to the first boundary of the drivable tunnel and the second boundary of the drivable tunnel, the drivable tunnel of the target path point.

[0035] Optionally, the generating, according to the center of the first disc, the inflated obstacle and the preset step length, the first boundary of the drivable tunnel comprises:

[0036] The generating, according to the center of the first disc, the inflated obstacle and the preset step length, the first boundary of the drivable tunnel comprises:

[0037] The determining whether the rectangle overlaps with the inflated obstacle.

[0038] If it is determined that the rectangle does not overlap with the inflated obstacle, combining all the rectangles obtained currently, and extending and constructing a rectangle in a preset step length to a next target direction according to an order of the target directions.

[0039] If it is determined that the rectangle overlaps with the inflated obstacle, removing the target direction from the order of the target directions, and continuing to extend and construct a rectangle in a preset step length to a next target direction according to the order of the target directions.

[0040] If all rectangles constructed in all target directions overlap with the expanded obstacles, then the first boundary of the drivable tunnel is generated based on all rectangles that do not overlap with the expanded obstacles.

[0041] Optionally, generating the second boundary of the drivable tunnel based on the center of the second disk, the expanded obstacle, and a preset step length includes:

[0042] Using the center of the second disk as the initial point, a rectangle is constructed by expanding it sequentially according to the four target directions with a preset step size;

[0043] Determine whether the rectangle overlaps with the expanded obstacle;

[0044] If it is determined that the rectangle does not overlap with the expanded obstacle, then all the rectangles obtained so far are combined and expanded to the next target direction in the order of the target directions with a preset step size to construct a rectangle.

[0045] If it is determined that the rectangle overlaps with the expanded obstacle, the target direction is removed from the sequence of the target directions, and the rectangle is expanded to the next target direction with a preset step size according to the sequence of the target directions.

[0046] If all rectangles constructed in all target directions overlap with the expanded obstacles, then the second boundary of the drivable tunnel is generated based on all rectangles that do not overlap with the expanded obstacles.

[0047] Optionally, if the obstacle at the target path point is a crossable obstacle, constructing a drivable tunnel at the target path point to obtain a drivable tunnel at the target path point includes:

[0048] The target path point is designated as the first undetermined point.

[0049] The first undetermined point is moved toward the target direction by a preset step length to obtain the second undetermined path point in the target direction;

[0050] Based on the second undetermined path point, construct four tire envelopes corresponding to the second undetermined path point;

[0051] Determine whether the four tire envelopes corresponding to the second undetermined path point overlap with obstacles;

[0052] If it is determined that the four tire envelopes corresponding to the second undetermined path point do not overlap with the obstacle, then the second undetermined path point is taken as the new first undetermined path point, and the process returns to the step of moving the first undetermined point toward the target direction by a preset step length to obtain the second undetermined path point in the target direction.

[0053] If it is determined that the four tire envelopes corresponding to the second undetermined path point overlap with the obstacle, then the first undetermined point is taken as the drivable tunnel boundary in the target direction.

[0054] Based on the drivable tunnel boundaries in all target directions, generate the drivable tunnels for the target path points.

[0055] Optionally, optimizing the initial path based on the drivable tunnels of all the target path points and preset constraints to obtain the final path includes:

[0056] Based on the feasible tunnels at the target path points, construct external environmental constraints.

[0057] Based on the coordinates of the rear axle center point, the front wheel deflection angle, the rate of change of the front wheel deflection angle, and the front and rear wheelbase, construct the dynamic constraints of the system.

[0058] Establish the first boundary value constraint condition based on the coordinates of the starting point and the ending point;

[0059] Based on the boundary information of the map boundary and the front wheel deflection angle information of the vehicle, a second boundary value constraint condition is constructed; wherein, the boundary information of the map includes the minimum and maximum values ​​of the horizontal and vertical coordinates of the map boundary, and the front wheel deflection angle information includes the upper and lower bounds of the front wheel deflection angle and the upper and lower bounds of the front wheel deflection angle transformation rate.

[0060] A cost function is constructed based on the initial path length, the front wheel yaw angle adjustment coefficient, and the front wheel yaw angle change rate adjustment coefficient.

[0061] Under the condition that the external environment constraints, system dynamic constraints, first boundary value constraints, and second boundary value constraints are satisfied, the path corresponding to the solution with the minimum cost function is taken as the final path.

[0062] A second aspect of this application provides a path planning apparatus, comprising:

[0063] The initial planning unit is used to perform preliminary path planning in the Frenet coordinate system of the reference path to obtain the initial path;

[0064] The first generation unit is used to generate a collision envelope for the vehicle for each obstacle in the initial path, based on the type of the obstacle.

[0065] The first judgment unit is used to determine whether the collision envelope of the vehicle on the initial path overlaps with the obstacle;

[0066] The first determining unit is configured to take the current path point as the target path point if the first determining unit determines that the collision envelope of the vehicle on the initial path overlaps with the obstacle.

[0067] A construction unit is used to construct a drivable tunnel for the target path point, thereby obtaining a drivable tunnel for the target path point;

[0068] An optimization unit is used to optimize the initial path based on the drivable tunnels of all the target path points and preset constraints to obtain the final path.

[0069] Optionally, the initial planning unit includes:

[0070] The first partitioning unit is used to divide the vertical axis of the Frenet coordinate system of the reference path into N sampling layers; where N is a positive integer.

[0071] The second partitioning unit is used to sample M sampling points at equal intervals along one boundary to the other boundary of the horizontal axis of the Frenet coordinate system of the reference path for each sampling layer; where M is a positive integer.

[0072] The cost calculation unit is used to calculate the cost of each sampling point in each sampling layer and record the parent node of the sampling point.

[0073] The backtracking unit is used to backtrack from the sampling point with the lowest cost in the last sampling layer, following the parent node of the sampling point to the previous sampling layer until the starting point is reached, so as to obtain the initial path.

[0074] Optionally, the cost calculation unit includes:

[0075] The cost calculation subunit is used to determine the cost of the sampling point based on the lateral movement cost, the change cost of lateral offset, the rate of change cost of lateral offset, and the collision cost.

[0076] Optionally, if there are crossable obstacles in the reference path, the path planning device further includes:

[0077] The second determining unit is used to determine the nearest sampling point of the crossable obstacle in the Frenet coordinate system of the reference path based on the coordinates of the crossable obstacle in the Frenet coordinate system of the reference path.

[0078] The replacement unit is used to replace the x-coordinate of the most recently sampled point with the x-coordinate of the crossable obstacle in the Frenet coordinate system of the reference path.

[0079] Optionally, the second determining unit includes:

[0080] The second determining subunit is used to determine the nearest sampling layer of the crossable obstacle based on the ordinate of the crossable obstacle in the Frenet coordinate system of the reference path;

[0081] A distance calculation unit is used to calculate the distance between each sampling point in the nearest sampling layer and the crossable obstacle;

[0082] The second determining subunit is further configured to take the sampling point closest to the traversable obstacle as the closest sampling point of the traversable obstacle in the Frenet coordinate system of the reference path.

[0083] Optionally, sampling thresholds are reserved on one side and the other side of the horizontal axis of the Frenet coordinate system of the reference path.

[0084] Optionally, the first generation unit includes:

[0085] The first generation subunit is used to generate four tire envelopes for each obstacle in the initial path, if the type of the obstacle is a crossable obstacle.

[0086] The first generation subunit is further configured to generate a vehicle envelope if the type of obstacle is a bypass obstacle.

[0087] Optionally, if the obstacle at the target path point is a bypass obstacle, the construction unit includes:

[0088] The second generation unit is used to generate a first disk and a second disk at the target path point based on the vehicle's size information and the coordinate information of the rear axle center of the vehicle; wherein the first disk and the second disk have the same radius;

[0089] An expansion unit is used to expand the obstacle according to the radii of the first disk and the second disk to obtain an expanded obstacle;

[0090] The third generation unit is used to generate the first boundary of the drivable tunnel based on the center of the first disk, the expanded obstacle, and the preset step length.

[0091] The fourth generation unit is used to generate the second boundary of the drivable tunnel based on the center of the second disk, the expanded obstacles, and the preset step length.

[0092] The fifth generation unit is used to generate a drivable tunnel for the target path point based on the first boundary and the second boundary of the drivable tunnel.

[0093] Optionally, the third generation unit includes:

[0094] The first expansion unit is used to expand and construct a rectangle with a preset step size, starting from the center of the first disk and following the order of the four target directions.

[0095] The second judgment unit is used to determine whether the rectangle overlaps with the expanded obstacle;

[0096] The first combining unit is used to combine all the currently obtained rectangles and expand them to the next target direction with a preset step size in the order of the target directions if the second judging unit determines that the rectangle does not overlap with the expanded obstacle.

[0097] The first elimination unit is used to eliminate the target direction from the order of the target directions if the second judgment unit determines that the rectangle overlaps with the expanded obstacle, and continue to expand and construct the rectangle in the next target direction with a preset step size according to the order of the target directions.

[0098] The sixth generation unit is used to generate the first boundary of the drivable tunnel based on all rectangles that do not overlap with the expanded obstacles when all rectangles constructed in all target directions overlap with the expanded obstacles.

[0099] Optionally, the fourth generation unit includes:

[0100] The second expansion unit is used to expand and construct a rectangle with a preset step size, starting from the center of the second disk and following the order of the four target directions.

[0101] The third judgment unit is used to determine whether the rectangle overlaps with the expanded obstacle;

[0102] The second combination unit is used to construct a rectangle by expanding the combination of all the rectangles obtained so far into the next target direction with a preset step size, according to the order of the target directions, if the third judgment unit determines that the rectangle does not overlap with the expanded obstacle.

[0103] The second elimination unit is used to eliminate the target direction from the order of the target directions if the third judgment unit determines that the rectangle overlaps with the expanded obstacle, and continue to expand and construct the rectangle in the next target direction with a preset step size according to the order of the target directions.

[0104] The seventh generation unit is used to generate the second boundary of the drivable tunnel based on all rectangles that do not overlap with the expanded obstacles when all rectangles constructed in all target directions overlap with the expanded obstacles.

[0105] Optionally, if the obstacle at the target path point is a crossable obstacle, the building unit includes:

[0106] The setting unit is used to designate the target path point as the first undetermined point;

[0107] A moving unit is used to move the first undetermined point toward the target direction by a preset step length to obtain a second undetermined path point in the target direction;

[0108] Construct sub-units to build four tire envelopes corresponding to the second undetermined path point based on the second undetermined path point;

[0109] The fourth judgment unit is used to determine whether the four tire envelopes corresponding to the second undetermined path point overlap with the obstacle;

[0110] An activation unit is used to activate the movement unit to execute the step of moving the first undetermined point towards the target direction by a preset step length to obtain the second undetermined path point if the fourth judgment unit determines that the four tire envelopes corresponding to the second undetermined path point do not overlap with the obstacle.

[0111] The third determining unit is used to determine the first undetermined point as the drivable tunnel boundary in the target direction if the fourth determining unit determines that the four tire envelopes corresponding to the second undetermined path point overlap with the obstacle.

[0112] The eighth generation unit is used to generate a drivable tunnel for the target path point based on the drivable tunnel boundaries in all target directions.

[0113] Optionally, the optimization unit includes:

[0114] The first constraint unit is used to construct external environmental constraint conditions based on the feasible tunnels of the target path point;

[0115] The second constraint unit is used to construct the dynamic constraint conditions of the system based on the coordinates of the rear axle center point of the vehicle, the front wheel deflection angle of the vehicle, the rate of change of the front wheel deflection angle, and the front and rear wheelbase.

[0116] The third constraint unit is used to establish the first boundary value constraint conditions based on the coordinates of the starting point and the ending point.

[0117] The fourth constraint unit is used to construct the second boundary value constraint condition based on the boundary information of the map boundary and the front wheel deflection angle information of the vehicle; wherein, the boundary information of the map includes the minimum and maximum values ​​of the horizontal and vertical coordinates of the map boundary, and the front wheel deflection angle information includes the upper and lower bounds of the front wheel deflection angle and the upper and lower bounds of the front wheel deflection angle transformation rate.

[0118] The cost function construction unit is used to construct the cost function based on the initial path length, the front wheel yaw angle adjustment coefficient, and the front wheel yaw angle change rate adjustment coefficient.

[0119] The fourth determining unit is used to select the path corresponding to the solution with the minimum cost function as the final path, provided that the external environment constraints, system dynamic constraints, first boundary value constraints, and second boundary value constraints are satisfied.

[0120] A third aspect of this application provides an electronic device, comprising:

[0121] One or more processors;

[0122] A storage device on which one or more programs are stored;

[0123] When the one or more programs are executed by the one or more processors, the one or more processors implement the path planning method as described in any one of the first aspects.

[0124] The fourth aspect of this application provides a computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the path planning method as described in any one of the first aspects.

[0125] As can be seen from the above solutions, this application provides a path planning method, apparatus, electronic device, and computer storage medium. The method involves preliminary path planning in the Frenet coordinate system of a reference path to obtain an initial path; for each obstacle in the initial path, a collision envelope of the vehicle is generated according to the type of the obstacle; if the collision envelope of the vehicle overlaps with the obstacle on the initial path, the current path point is taken as the target path point; then, a drivable tunnel is constructed for the target path point to obtain a drivable tunnel; finally, the initial path is optimized based on the drivable tunnels of all target path points and preset constraints to obtain the final path. This achieves the goal of efficiently and stably planning obstacle avoidance paths when facing different obstacle types. Attached Figure Description

[0126] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0127] Figure 1 A detailed flowchart of a path planning method provided in an embodiment of this application;

[0128] Figure 2 A flowchart illustrating an initial path planning method provided in another embodiment of this application;

[0129] Figure 3 A schematic diagram illustrating a method for generating sampling points according to another embodiment of this application;

[0130] Figure 4 A flowchart of a method for correcting the coordinates of sampling points when there are crossable obstacles in the Frenet coordinate system of the reference path, provided in another embodiment of this application;

[0131] Figure 5 A flowchart illustrating a method for determining the nearest sampling point, provided in another embodiment of this application;

[0132] Figure 6 A schematic diagram illustrating a method for correcting the coordinates of sampling points when there are crossable obstacles in the Frenet coordinate system of the reference path, as provided in another embodiment of this application;

[0133] Figure 7 A flowchart of a method for constructing a drivable tunnel when the type of obstacle at the target path point is a bypass obstacle, provided in another embodiment of this application;

[0134] Figure 8 A schematic diagram of a circular mass replacing a vehicle, provided for another embodiment of this application;

[0135] Figure 9 A flowchart illustrating a method for generating the first boundary of a drivable tunnel, as provided in another embodiment of this application;

[0136] Figure 10 A schematic diagram illustrating the generation of a first boundary of a drivable tunnel, as provided in another embodiment of this application;

[0137] Figure 11 A flowchart of a method for constructing a drivable tunnel when the obstacle at the target path point is a crossable obstacle, provided in another embodiment of this application;

[0138] Figure 12 A schematic diagram illustrating the construction of a drivable tunnel when the obstacle at the target path point is a crossable obstacle, as provided in another embodiment of this application;

[0139] Figure 13 A flowchart illustrating a method for optimizing an initial path to obtain a final path, as provided in another embodiment of this application;

[0140] Figure 14A schematic diagram of a two-degree-of-freedom vehicle kinematic model provided for another embodiment of this application;

[0141] Figure 15 A schematic diagram of a path planning device provided in another embodiment of this application;

[0142] Figure 16 This is a schematic diagram of an electronic device that implements a path planning method, as provided in another embodiment of this application. Detailed Implementation

[0143] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0144] The term "comprising" and its variations as used herein are open-ended inclusions, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.

[0145] It should be noted that the concepts of "first" and "second" mentioned in this application are only used to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0146] It should be noted that the terms "a" and "a plurality of" used in this application are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0147] This application provides a path planning method, such as... Figure 1 As shown, the specific steps include:

[0148] S101. Perform preliminary path planning in the Frenet coordinate system of the reference path to obtain the initial path.

[0149] The reference path can be a fixed driving route, historical path, etc., and is not limited here. The Frenet coordinate system is a commonly used coordinate system in the field of autonomous driving. It is based on the center line of the road as the reference line and establishes the coordinate system through tangent vectors and normal vectors. In the Frenet coordinate system, the position of the vehicle is described by two coordinates, s and d. s represents the longitudinal displacement along the road, also known as the ordinate, while d represents the lateral displacement, that is, the distance the vehicle deviates from the center of the road, also known as the abscissa.

[0150] Optionally, in another embodiment of this application, one implementation of step S101 is as follows: Figure 2 As shown, it includes:

[0151] S201. Divide the vertical axis of the Frenet coordinate system of the reference path into N equal sampling layers.

[0152] Where N is a positive integer.

[0153] In the specific implementation process of this application, you can refer to Figure 3 Where S0 is the ordinate of the starting point of the reference path, and S0+S max It is the ordinate of the endpoint of the reference path.

[0154] S202. For each sampling layer, M sampling points are sampled at equal intervals from one boundary to the other boundary along the horizontal axis of the Frenet coordinate system of the reference path on the sampling layer.

[0155] Where M is a positive integer.

[0156] Continuing with the above example, with M=5 and taking the first layer as an example, five sampling points are sampled at equal intervals from one side boundary (such as the left edge of the road) to the other side boundary (such as the right edge of the road) of the horizontal axis of the Frenet coordinate system of the reference path.

[0157] It is understandable that, since the distance from one boundary (such as the left edge of the road) to the other boundary (such as the right edge of the road) of the horizontal axis may be different in different sampling layers of the Frenet coordinate system of the reference path, the interval between two sampling points in different sampling layers may be different.

[0158] It should be noted that in the actual application of this application, in order to improve the efficiency of obstacle avoidance path planning, the number of sampling points should not be too large. The number of sampling points is the product of M and N, that is, the number of sampling layers and the number of sampling points in each layer should not be too large.

[0159] Furthermore, considering the safety of the obstacle avoidance process, sampling thresholds can be reserved on one and the other sides of the horizontal axis of the Frenet coordinate system of the reference path. This ensures that sampling points are not directly on the left and right boundaries. The size of the sampling threshold can be, but is not limited to, half the vehicle width multiplied by a safety factor; no limitation is imposed here. The safety factor is preset and modified by experts or authorized technicians, and no limitation is imposed here either.

[0160] S203. Calculate the cost of each sampling point in each sampling layer and record the parent node of the sampling point.

[0161] In the specific implementation of this application, the cost of sampling points can be determined by, but is not limited to, the cost of lateral movement of sampling points, the cost of lateral offset variation, the cost of lateral offset change rate, and the cost of collision. No limitation is made here.

[0162] It should be noted that the order in which the cost of the sampling points is calculated can be, but is not limited to, from left to right or from top to bottom, and the parent node of the sampling point is recorded. There is no restriction here.

[0163] The specific formula for calculating the cost of a sampling node can be as follows:

[0164] C total =C d +C d_change +C d_rate +C collision ;

[0165] Among them, C total C represents the cost of a sampling node. d C represents the cost of lateral offset. d_change C represents the cost of changes in lateral offset. d_rate C represents the cost of the rate of change of lateral offset. collision Let represent the collision cost. The lateral offset cost is the d-value of the sampling point in the Frenet coordinate system. The cost of the change in lateral offset is the first derivative of the lateral offset, and the cost of the rate of change of lateral offset is the second derivative of the lateral offset. The collision cost can be calculated as follows: if there is a collision, assign a large value (1000); if there is no collision, calculate the collision cost as e based on the distance between the sampling point and the bypass obstacle. -d / ω , where d is the Euclidean distance and ω is a constant coefficient (preset).

[0166] Of course, in the specific implementation of this application, weighted values ​​can be set for each parameter (lateral offset cost, lateral offset change cost, lateral offset change rate cost, collision cost), and the calculation formula for the cost of the sampling node can be as follows:

[0167] C total =ωd ·C d +ω d_change ·C d_change +ω d_rate ·C d_rate +ω collision ·C collision ;

[0168] Where, ω d ω is the weighted value of the lateral offset cost. d_change ω is the weighted value of the cost of the change in lateral offset. d_rate ω is the weighted value of the cost of the rate of change of the lateral offset. collision This represents the weighted value for collision costs. The weighted value for each parameter is a value obtained by experts and technicians based on experience and continuous adjustments; no specific limit is imposed here.

[0169] In the specific implementation of this application, the parent node records the path from the starting point to the current node. By recording the parent node information, a complete path from the starting point to the destination can be easily constructed. Furthermore, during the search for the optimal path, if it is found that the path to the destination through a certain node is not optimal, it is possible to backtrack to previous nodes through its parent node and try other paths, thereby finding a better solution and avoiding the repeated calculation of the same path during the search for the optimal solution, effectively improving efficiency.

[0170] S204. Starting from the sampling point with the lowest cost in the last sampling layer, backtrack sequentially according to the parent node of the sampling point to the previous sampling layer until the starting point is reached to obtain the initial path.

[0171] like Figure 3 As shown, assume the Nth S The third sampling point of the layer is N S The sampling point with the lowest cost in the layer is then selected based on the information recorded in the parent node of this sampling point. S The sampling point with the lowest cost in layer -1 is selected, and so on, until the starting point is traced back to obtain the initial path.

[0172] In the specific implementation of this application, if there are crossable obstacles in the reference path, to prevent collisions with these obstacles due to errors in control and positioning, the final obstacle avoidance path should be made as close as possible to the crossable obstacles. Therefore, in another embodiment of this application, after sampling M sampling points at equal intervals along one boundary to the other boundary of the Frenet coordinate system of the reference path for each sampling layer, one implementation of the path planning method is as follows: Figure 4 As shown, it also includes:

[0173] S401. Based on the coordinates of the crossable obstacle in the Frenet coordinate system of the reference path, determine the nearest sampling point of the crossable obstacle in the Frenet coordinate system of the reference path.

[0174] First, in the specific implementation process of this application, the type of obstacle is determined based on the perception and recognition results. For example, if the length, width and height of the obstacle are all less than 0.2m, the obstacle is determined to be a crossable obstacle. If any of the length, width or height is greater than 0.2m, the obstacle is determined to be a detour obstacle. Or, if the obstacle is not made of rock or other materials, it is determined to be a crossable obstacle. If the obstacle is made of rock, it is determined to be a detour obstacle. No limitation is made here.

[0175] When performing object perception and recognition, deep learning models can be used, but are not limited to, to perform object perception and recognition; no limitation is made here.

[0176] For example, Convolutional Neural Networks (CNNs) are widely used for predicting and classifying the materials of objects. By training on a large amount of material image data, they can learn the visual features of different materials, thereby automatically identifying and classifying the materials in new images; this is not a limitation.

[0177] Optionally, in another embodiment of this application, one implementation of step S401 is as follows: Figure 5 As shown, it includes:

[0178] S501. Determine the nearest sampling layer of the crossable obstacle based on the ordinate of the crossable obstacle in the Frenet coordinate system of the reference path.

[0179] Specifically, the sampling layer closest to the crossable obstacle is determined based on the ordinate of the crossable obstacle in the Frenet coordinate system of the reference path and the ordinate of the sampling layer, and is taken as the nearest sampling layer.

[0180] S502. For each sampling point in the nearest sampling layer, calculate the distance between the sampling point and the crossable obstacle.

[0181] Specifically, for each sampling point in the most recent sampling layer, the difference between the horizontal coordinate of the sampling point and the horizontal coordinate of the crossable obstacle is calculated, and this difference is used as the distance between the sampling point and the crossable obstacle.

[0182] S503. The sampling point closest to the crossable obstacle is taken as the closest sampling point of the crossable obstacle in the Frenet coordinate system of the reference path.

[0183] S402. Replace the x-coordinate of the most recently sampled point with the x-coordinate of the crossable obstacle in the Frenet coordinate system of the reference path.

[0184] like Figure 6As shown, for a crossable obstacle (S0, d0), find its nearest sampling point (S0, d0) on the d-axis. S d S ), and set the x-coordinate d of the most recent sampled point. S Replace the x-coordinates of the crossable obstacles in the Frenet coordinate system of the reference path to obtain the new sampling point coordinates (S). S This ensures that the planned initial path falls on the crossable obstacle, avoiding collisions with it.

[0185] S102. For each obstacle in the initial path, generate the collision envelope of the vehicle according to the type of obstacle.

[0186] In the specific implementation of this application, for each obstacle in the initial path, if the obstacle is a crossable obstacle, four tire envelopes are generated; if the obstacle is a bypass obstacle, a whole vehicle envelope is generated.

[0187] Specifically, the method for generating the collision envelope of a vehicle can be, but is not limited to, using ADAMS software for simulation modeling; no limitation is made here.

[0188] S103. Determine whether the collision envelope of the vehicle overlaps with the obstacle on the initial path.

[0189] Understandably, when the collision envelope of a vehicle is defined as the envelope of the four tires, if the tire envelope overlaps with a crossable obstacle, it indicates that the tires collided with the crossable obstacle. When the collision envelope of a vehicle is defined as the entire vehicle envelope, if the entire vehicle envelope overlaps with a bypass obstacle, it indicates that the vehicle collided with the bypass obstacle.

[0190] Specifically, if it is determined that the collision envelope of the vehicle overlaps with the obstacle on the initial path, then step S104 is executed.

[0191] S104. Set the current path point as the target path point.

[0192] Continuing with the above examples, we can see that target waypoints can be divided into detour waypoints and crossable waypoints.

[0193] S105. Construct a drivable tunnel for the target path point to obtain the drivable tunnel for the target path point.

[0194] Since target path points are divided into detour path points and crossable path points, different methods for constructing drivable tunnels will be adopted for different types of path points in the specific implementation of this application.

[0195] Optionally, in another embodiment of this application, if the obstacle at the target path point is a bypass obstacle, i.e., the target path point is a bypass path point, one implementation of step S105 is as follows: Figure 7 As shown, it includes:

[0196] S701. At the target path point, generate the first disk and the second disk based on the vehicle's size information and the coordinate information of the rear axle center.

[0197] The first and second disks have the same radius.

[0198] It should be noted that the vehicle's dimensional information includes, but is not limited to, the coordinates (horizontal and vertical coordinates) of the vehicle's rear axle center in the Frenet coordinate system of the reference path, the front and rear wheelbases, the front overhang distance, the vehicle width, and the vehicle's heading angle, etc., which are not limited here.

[0199] In the specific implementation of this application, the coordinates of the center of the first disk and the center of the second disk can be determined according to the following formula:

[0200]

[0201] Where x represents the x-coordinate of the rear axle center of the vehicle; y represents the y-coordinate of the rear axle center of the vehicle; x f The x-coordinate of the center of the first disk is represented by y. f The x-coordinate represents the ordinate of the center of the first disk. r The x-coordinate of the center of the second disk is represented by y. r L represents the x-coordinate of the center of the second disk. w Indicates the wheelbase between the front and rear wheels; L F Indicates the front overhang distance; L R L represents the rear overhang distance. B θ represents the vehicle width; θ represents the vehicle heading angle.

[0202] Furthermore, it can also be based on L w L F L R L B The radii R of the first and second disks were calculated as follows:

[0203]

[0204] Thus, the first disk can be generated in the Frenet coordinate system of the reference path based on the center coordinates and radius of the first disk, and the second disk can be generated in the Frenet coordinate system of the reference path based on the center coordinates and radius of the second disk.

[0205] S702. The obstacle is expanded according to the radii of the first and second disks to obtain the expanded obstacle.

[0206] In the specific implementation of this application, the radius of the first disk can be increased to expand the obstacle, resulting in an expanded obstacle.

[0207] Understandably, after expanding the obstacle, it's only necessary to ensure that the centers of the two disks remain aligned so that the expanded obstacle doesn't collide. For example... Figure 8 The diagram shown is a schematic representation of a circular mass replacing a vehicle, as provided in this application.

[0208] It should be noted that, since the initial path may cross map boundaries, in the practical application of this application, when the initial path crosses a map boundary, the map boundary should be contracted, that is, the map boundary should be reduced by the radius of the first disk. This is to avoid collisions between the vehicle and the map boundary.

[0209] S703. Generate the first boundary of the drivable tunnel based on the center of the first disk, the expanded obstacle, and the preset step length.

[0210] The preset step size is an empirical value obtained by experts and technicians based on accuracy requirements and construction time, and is not limited here.

[0211] Optionally, in another embodiment of this application, one implementation of step S703 is as follows: Figure 9 As shown, it includes:

[0212] S901. Starting from the center of the first disk, construct a rectangle by expanding it sequentially according to the four target directions with a preset step size.

[0213] In the specific implementation of this application, the four target directions can be, but are not limited to, due north, due south, due east, and due west; no limitation is made here.

[0214] like Figure 10 As shown, the center of the first disk is the center point (x0, y0) in the diagram. A rectangle is constructed by expanding it in the directions of west, north, east, and south with a preset step size Δs. First, the center point is moved westward by a step size Δs to obtain (x0-Δs, y0). (x0-Δs, y0) forms one side of the rectangle with the center point (x0, y0). Then, it is moved northward by a step size Δs to obtain (x0-Δs, y0+Δs). (x0-Δs, y0+Δs) forms the other side of the rectangle with (x0-Δs, y0). This completes the construction of the rectangle. Figure 10 Rectangle 2 in the middle.

[0215] S902. Determine whether the rectangle overlaps with the expanded obstacle.

[0216] Specifically, if it is determined that the rectangle does not overlap with the expanded obstacle, then step S903 is executed; if it is determined that the rectangle overlaps with the expanded obstacle, then step S904 is executed.

[0217] S903: Combine all the rectangles obtained so far.

[0218] Continuing with the above example, if rectangle 2 does not overlap with the expanded obstacle, since there is only a center point initially and no rectangle, rectangle 3 is constructed by expanding eastward with a preset step size Δs, using rectangle 2 as the reference. If rectangle 3 also does not overlap with the expanded obstacle, rectangle 2 and rectangle 3 are combined to obtain combined rectangle 23. Using combined rectangle 23 as the reference, rectangle 4 is constructed by expanding southward with a preset step size. If rectangle 4 also does not overlap with the expanded obstacle, rectangle 2, rectangle 3, and rectangle 4 are combined to obtain combined rectangle 234. Using combined rectangle 234 as the reference, rectangle 5 is constructed by expanding westward with a preset step size, and so on.

[0219] S904. Eliminate the target direction from the order of target directions.

[0220] S905. Construct a rectangle by expanding it in the next target direction with a preset step size, following the order of the target directions.

[0221] Continuing with the example above, after obtaining rectangle 5, rectangles 2, 3, 4, and 5 are combined to obtain combined rectangle 2345. Using combined rectangle 2345 as the reference, a rectangle is constructed by expanding it in the due north direction with a preset step size to obtain rectangle 6. At this point, rectangle 6 overlaps with the expanded obstacle. Therefore, the due north direction is removed from the order of the target directions. In other words, in the subsequent process of expanding the rectangle in the order of the target directions with a preset step size, it will no longer expand in the due north direction because the generated rectangle 6 has already overlapped with the expanded obstacle. Subsequent expansion in the due north direction to construct a rectangle will also overlap with the obstacle.

[0222] After eliminating the north direction from the order of the target directions, continue to use the combined rectangle 2345 as the reference, and expand the rectangle to the next target direction, namely the east direction, with a preset step size to obtain rectangle 7.

[0223] S906. When all rectangles constructed in all target directions overlap with the expanded obstacles, the first boundary of the drivable tunnel is generated based on all rectangles that do not overlap with the expanded obstacles.

[0224] Continuing with the previous example, rectangles 2, 3, 4, 5, and 7 are combined to obtain combined rectangle 23457. Using combined rectangle 23457 as a reference, a rectangle is constructed by expanding southward with a preset step size, resulting in rectangle 8. If rectangle 8 does not overlap with the expanded obstacle, then rectangles 2, 3, 4, 5, 7, and 8 are combined to obtain combined rectangle 234578. Using combined rectangle 234578 as a reference, a rectangle is constructed by expanding westward with a preset step size, resulting in rectangle 9. If rectangle 9 overlaps with the expanded obstacle, the westward direction is removed from the target direction order. This process continues until all rectangles constructed in all target directions overlap with the expanded obstacle, meaning there are no more target directions to expand and construct rectangles with a preset step size. At this point, all rectangles that do not overlap with the expanded obstacle are combined to obtain the final determined rectangular area. The boundary of this final determined rectangular area is used as the first boundary of the drivable tunnel. Figure 10 For example, the combined rectangle is 23457811.

[0225] It should be noted that in the specific implementation of this application, the expansion of the constructed rectangle can be stopped after collisions occur in all four target directions as described in the above embodiment, or an upper limit for the number of expansions can be preset, and the expansion of the constructed rectangle can be stopped when the upper limit for the number of expansions is reached. No limitation is made here.

[0226] S704. Generate the second boundary of the drivable tunnel based on the center of the second disk, the expanded obstacles, and the preset step length.

[0227] It should be noted that the specific implementation method of step S704 can be found in the specific implementation method of step S703, and will not be repeated here.

[0228] S705. Generate a drivable tunnel for the target path point based on the first boundary and the second boundary of the drivable tunnel.

[0229] Specifically, the first boundary and the second boundary of the drivable tunnel are combined to obtain the drivable tunnel for the target path point.

[0230] Optionally, in another embodiment of this application, if the obstacle at the target path point is a crossable obstacle, i.e., the target path point is a crossable path point, one implementation of step S105 is as follows: Figure 11 As shown, it includes:

[0231] S1101, designate the target path point as the first undetermined point.

[0232] S1102. Move the first undetermined point toward the target direction by a preset step length to obtain the second undetermined path point in the target direction.

[0233] like Figure 12 The diagram shows the construction of a drivable tunnel when the target path point is a crossable path point; it includes four target directions, such as direction 1, direction 2, direction 3, and direction 4. Figure 12 In this context, p0 is the target path point. The target path point is taken as the first undetermined point. Taking direction 1 as an example, the first undetermined point is moved by a preset step length Δs in direction 1 to obtain P1, which is the second undetermined path point in direction 1.

[0234] S1103. Based on the second undetermined path point, construct the four tire envelopes corresponding to the second undetermined path point.

[0235] Specifically, the specific implementation method for constructing the tire envelope can be, but is not limited to, using ADAMS software for simulation modeling; no limitation is made here.

[0236] S1104. Determine whether the four tire envelopes corresponding to the second undetermined path point overlap with obstacles.

[0237] Specifically, if it is determined that the four tire envelopes corresponding to the second undetermined path point do not overlap with the obstacle, then step S1105 is executed; if it is determined that the four tire envelopes corresponding to the second undetermined path point overlap with the obstacle, then step S1106 is executed.

[0238] S1105. Take the second undetermined path point as the new first undetermined path point, and return to execute the step of moving the first undetermined point towards the target direction by a preset step length to obtain the second undetermined path point in the target direction.

[0239] S1106. Take the first undetermined point as the boundary of the drivable tunnel in the target direction.

[0240] Continuing with the above example, based on the second undetermined path point, construct the four tire envelopes corresponding to the second undetermined path point. If it is determined that the four tire envelopes corresponding to the second undetermined path point do not overlap with the obstacle, then the second undetermined path point P2 is taken as the new first undetermined path point P1, and step S1102 is executed again. If it is determined that the four tire envelopes corresponding to the second undetermined path point overlap with the obstacle, then the first undetermined point is taken as the drivable tunnel boundary in the target direction.

[0241] S1107. Generate drivable tunnels for target path points based on the drivable tunnel boundaries in all target directions.

[0242] Specifically, when all target directions overlap with obstacles, a drivable tunnel for the target path point is generated based on all determined drivable tunnel boundaries.

[0243] S106. Optimize the initial path based on the drivable tunnels of all target path points and the preset constraints to obtain the final path.

[0244] In the specific implementation of this application, the preset constraints include, but are not limited to, external environment constraints, system dynamic constraints, boundary value constraints, etc., which are not limited here.

[0245] Optionally, in another embodiment of this application, one implementation of step S106 is as follows: Figure 13 The following are included:

[0246] S1301. Based on the feasible tunnels of the target path points, construct external environmental constraints.

[0247] In the specific implementation of this application, if the obstacle is a bypass obstacle, then based on the drivable tunnel of the target path point, external environmental constraint conditions are constructed, including: generating constraint conditions for the first and second circles based on the first and second boundaries of the drivable tunnel.

[0248]

[0249] Where, x fmin (s) represents the minimum x-coordinate of the center of the first disk; x fmax (s) represents the maximum x-coordinate of the center of the first disk; x rmin (s) represents the minimum x-coordinate of the center of the second disk; x rmax (s) represents the maximum value of the x-coordinate of the center of the second disk; y fmin (s) represents the minimum value of the ordinate of the center of the first disk; y fmax (s) represents the maximum value of the ordinate of the center of the first disk; y rmin (s) represents the minimum value of the ordinate of the center of the second disk; y rmax (s) represents the maximum value of the ordinate of the center of the second disk. f (s) is the x-coordinate of the center of the first disk, y f (s) is the ordinate of the center of the first disk, x r (s) is the x-coordinate of the center of the second disk, y r (s) is the ordinate of the center of the second disk.

[0250] In the specific implementation of this application, if the obstacle is a traversable obstacle, then based on the drivable tunnel of the target path point, external environmental constraint conditions are constructed, including: generating constraint conditions for the target path point based on the boundary of the drivable tunnel.

[0251]

[0252] Where, x c_min x represents the minimum x-coordinate of the target path point. c_max The x-coordinate of the target path point is represented by the maximum value of y. c_min The minimum value of the y-coordinate of the target path point. c_max y(s) represents the maximum value of the ordinate of the target path point, x(s) represents the abscissa of the target path point, and y(s) represents the ordinate of the target path point.

[0253] S1302. Based on the coordinates of the rear axle center point, the front wheel deflection angle, the rate of change of the front wheel deflection angle, and the front and rear wheelbase, construct the dynamic constraints of the system.

[0254] For path planning problems, using two-degree-of-freedom vehicle kinematic constraints as system dynamic constraints is sufficient to meet practical requirements. Figure 14 In the kinematic model, the vehicle's motion in the Cartesian coordinate system is constrained by the following differential equations, i.e., the system dynamic constraints:

[0255]

[0256] Where (x(s), y(s)) represents the center point of the rear axle of the vehicle; θ(s) represents the coordinate vehicle attitude angle; φ(s) represents the front wheel yaw angle of the vehicle; ω(s) represents the rate of change of the front wheel yaw angle, and L w This indicates the wheelbase between the front and rear wheels.

[0257] S1303. Establish the first boundary value constraint condition based on the coordinates of the starting point and the ending point.

[0258] Specifically, the start and end point constraints are used to limit the vehicle's motion state at the start and end times to ensure the curvature continuity of the final obstacle avoidance path. The first boundary constraint can be described as follows:

[0259]

[0260] Where, x s The x-coordinate of the starting point, y s The ordinate of the starting point, x g The x-coordinate of the endpoint is represented by y. g Represents the ordinate of the endpoint.

[0261] S1304. Based on the boundary information of the map boundary and the vehicle front wheel deflection angle information, construct the second boundary value constraint condition.

[0262] The map boundary information includes the minimum and maximum values ​​of the horizontal and vertical coordinates of the map boundary, and the front wheel deflection angle information includes the upper and lower bounds of the front wheel deflection angle and the upper and lower bounds of the front wheel deflection angle transformation rate.

[0263] It should be noted that the map boundary, vehicle mechanical characteristics, etc., determine the allowable range of state variables. The second boundary value constraint is as follows:

[0264]

[0265] Where, x min The x-coordinate of the map boundary is the minimum value; y min The minimum ordinate of the map boundary; x max The x-coordinate represents the maximum value of the map boundary; y max φ represents the maximum value of the ordinate of the map boundary. min Indicates the upper limit of the front wheel deflection angle; φ max Indicates the lower limit of the front wheel deflection angle; ω min This represents the upper bound of the rate of change of the front wheel yaw angle; ω max This indicates the lower bound of the rate of change of the front wheel yaw angle.

[0266] S1305. Construct a cost function based on the initial path length, the front wheel yaw angle adjustment coefficient, and the front wheel yaw angle change rate adjustment coefficient.

[0267] In the specific implementation of this application, the cost function J can be as follows:

[0268] J = s f +w phi ∫[φ(s)] 2 ds+w omega ∫[ω(s)] 2 ds

[0269] Among them, s f Indicates the initial path length; w phi Indicates the front wheel yaw angle adjustment coefficient; w omega This represents the adjustment coefficient for the rate of change of the front wheel yaw angle.

[0270] Understandably, the first term of the cost function guarantees maximum operating efficiency, the second term indicates that the front wheel yaw angle should be as small as possible, and the third term indicates that the rate of change of the front wheel yaw angle should be as small as possible.

[0271] S1306. Under the condition of satisfying the external environment constraint conditions, system dynamic constraint conditions, first boundary value constraint conditions and second boundary value constraint conditions, the path corresponding to the solution with the minimum cost function shall be taken as the final path.

[0272] As can be seen from the above scheme, this application provides a path planning method. After obtaining an initial path through preliminary path planning in the Frenet coordinate system of the reference path, for each obstacle in the initial path, a collision envelope of the vehicle is generated according to the obstacle type. If the collision envelope of the vehicle overlaps with an obstacle on the initial path, the current path point is taken as the target path point. Then, a drivable tunnel is constructed for the target path point, resulting in a drivable tunnel for the target path point. Finally, the initial path is optimized based on the drivable tunnels of all target path points and preset constraints to obtain the final path. This achieves the goal of efficiently and stably planning obstacle avoidance paths when facing different obstacle types.

[0273] This application provides a path planning device, such as... Figure 15 As shown, it specifically includes:

[0274] Initial planning unit 1501 is used to perform preliminary path planning in the Frenet coordinate system of the reference path to obtain the initial path.

[0275] Optionally, in another embodiment of this application, one implementation of the initial planning unit 1501 includes:

[0276] The first partitioning unit is used to divide the vertical axis of the Frenet coordinate system of the reference path into N sampling layers.

[0277] Where N is a positive integer.

[0278] The second partitioning unit is used to sample M sampling points at equal intervals along one boundary to the other boundary of the horizontal axis of the Frenet coordinate system of the reference path for each sampling layer.

[0279] Where M is a positive integer.

[0280] The cost calculation unit is used to calculate the cost of each sampling point in each sampling layer and record the parent node of the sampling point.

[0281] The backtracking unit is used to backtrack from the sampling point with the lowest cost in the last sampling layer, following the parent node of the sampling point to the previous sampling layer until the starting point is reached, so as to obtain the initial path.

[0282] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, such as... Figure 2 As shown, it will not be elaborated further here.

[0283] Optionally, in another embodiment of this application, one implementation of the cost calculation unit includes:

[0284] The cost calculation subunit is used to determine the cost of a sampling point based on the lateral movement cost, the change cost of lateral offset, the rate of change cost of lateral offset, and the collision cost.

[0285] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, which will not be repeated here.

[0286] Optionally, in another embodiment of this application, if there are crossable obstacles in the reference path, one implementation of the path planning device further includes:

[0287] The second determining unit is used to determine the nearest sampling point of the crossable obstacle in the Frenet coordinate system of the reference path based on the coordinates of the crossable obstacle in the Frenet coordinate system of the reference path.

[0288] Replacement unit, used to replace the x-coordinate of the most recently sampled point with the x-coordinate of the crossable obstacle in the Frenet coordinate system of the reference path.

[0289] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, such as... Figure 4 As shown, it will not be elaborated further here.

[0290] Optionally, in another embodiment of this application, one implementation of the second determining unit includes:

[0291] The second determining subunit is used to determine the nearest sampling layer of the crossable obstacle based on the ordinate of the crossable obstacle in the Frenet coordinate system of the reference path.

[0292] The distance calculation unit is used to calculate the distance between each sampling point in the nearest sampling layer and the crossable obstacle.

[0293] The second determining sub-unit is also used to take the sampling point closest to the crossable obstacle as the nearest sampling point of the crossable obstacle in the Frenet coordinate system of the reference path.

[0294] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, such as... Figure 5 As shown, it will not be elaborated further here.

[0295] Optionally, sampling thresholds are reserved on one side and the other side of the horizontal axis of the Frenet coordinate system of the reference path.

[0296] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, which will not be repeated here.

[0297] The first generation unit 1502 is used to generate a collision envelope for the vehicle for each obstacle in the initial path, based on the type of obstacle.

[0298] Optionally, in another embodiment of this application, one implementation of the first generating unit 1502 includes:

[0299] The first generation subunit is used to generate four tire envelopes for each obstacle in the initial path, if the obstacle is a crossable obstacle.

[0300] The first generation subunit is also used to generate the whole vehicle envelope if the obstacle type is a bypass obstacle.

[0301] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, which will not be repeated here.

[0302] The first judgment unit 1503 is used to determine whether the collision envelope of the vehicle overlaps with the obstacle on the initial path.

[0303] The first determining unit 1504 is used to take the current path point as the target path point if the first determining unit determines that the collision envelope of the vehicle overlaps with the obstacle on the initial path.

[0304] Construction unit 1505 is used to construct a drivable tunnel for the target path point, thereby obtaining the drivable tunnel for the target path point.

[0305] Optionally, in another embodiment of this application, if the obstacle at the target path point is a bypass obstacle, one implementation of the construction unit 1505 includes:

[0306] The second generation unit is used to generate a first disk and a second disk at the target path point based on the vehicle's size information and the coordinate information of the vehicle's rear axle center.

[0307] The first and second disks have the same radius.

[0308] The expansion unit is used to expand the obstacle according to the radius of the first disk and the second disk to obtain the expanded obstacle.

[0309] The third generation unit is used to generate the first boundary of the drivable tunnel based on the center of the first disk, the expanded obstacles, and the preset step length.

[0310] The fourth generation unit is used to generate the second boundary of the drivable tunnel based on the center of the second disk, the expanded obstacles, and the preset step length.

[0311] The fifth generation unit is used to generate a drivable tunnel for the target path point based on the first boundary and the second boundary of the drivable tunnel.

[0312] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, such as... Figure 7 As shown, it will not be elaborated further here.

[0313] Optionally, in another embodiment of this application, one implementation of the third generation unit includes:

[0314] The first expansion unit is used to expand and construct a rectangle with a preset step size, starting from the center of the first disk and following the order of the four target directions.

[0315] The second judgment unit is used to determine whether the rectangle overlaps with the expanded obstacle.

[0316] The first combining unit is used to combine all the rectangles obtained so far and expand them to the next target direction with a preset step size in the order of the target directions if the second judgment unit determines that the rectangle does not overlap with the expanded obstacle.

[0317] The first elimination unit is used to eliminate the target direction in the order of the target directions if the second judgment unit determines that the rectangle overlaps with the expanded obstacle, and continue to expand and construct the rectangle in the next target direction with a preset step size in the order of the target directions.

[0318] The sixth generation unit is used to generate the first boundary of the drivable tunnel based on all rectangles that do not overlap with the expanded obstacles when all rectangles constructed in all target directions overlap with the expanded obstacles.

[0319] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, such as... Figure 9 As shown, it will not be elaborated further here.

[0320] Optionally, in another embodiment of this application, one implementation of the fourth generation unit includes:

[0321] The second extension unit is used to construct a rectangle by extending it sequentially in four target directions with a preset step size, starting from the center of the second disk.

[0322] The third judgment unit is used to determine whether the rectangle overlaps with the expanded obstacle.

[0323] The second combining unit is used to combine all the currently obtained rectangles and expand them to the next target direction with a preset step size in the order of the target directions if the third judgment unit determines that the rectangle does not overlap with the expanded obstacle.

[0324] The second elimination unit is used to eliminate the target direction in the order of the target directions if the third judgment unit determines that the rectangle overlaps with the expanded obstacle, and continue to expand and construct the rectangle in the next target direction with a preset step size in the order of the target directions.

[0325] The seventh generation unit is used to generate the second boundary of the drivable tunnel based on all rectangles that do not overlap with the expanded obstacles when all rectangles constructed in all target directions overlap with the expanded obstacles.

[0326] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, which will not be repeated here.

[0327] Optionally, in another embodiment of this application, if the obstacle at the target path point is a crossable obstacle, one implementation of the construction unit 1505 includes:

[0328] The setting unit is used to designate the target path point as the first undetermined point.

[0329] The moving unit is used to move the first undetermined point in the target direction by a preset step length to obtain the second undetermined path point in the target direction.

[0330] Construct sub-units to build four tire envelopes corresponding to the second undetermined path point.

[0331] The fourth judgment unit is used to determine whether the four tire envelopes corresponding to the second undetermined path point overlap with obstacles.

[0332] The activation unit is used to activate the movement unit to execute the step of moving the first undetermined point towards the target direction by a preset step length if the fourth judgment unit determines that the four tire envelopes corresponding to the second undetermined path point do not overlap with the obstacle.

[0333] The third determining unit is used to determine the first undetermined point as the drivable tunnel boundary in the target direction if the fourth determining unit determines that the four tire envelopes corresponding to the second undetermined path point overlap with the obstacle.

[0334] The eighth generation unit is used to generate drivable tunnels for target path points based on the drivable tunnel boundaries in all target directions.

[0335] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, such as... Figure 11 As shown, it will not be elaborated further here.

[0336] The optimization unit 1506 is used to optimize the initial path based on the drivable tunnels of all target path points and preset constraints to obtain the final path.

[0337] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, such as... Figure 1 As shown, it will not be elaborated further here.

[0338] Optionally, in another embodiment of this application, one implementation of the optimization unit 1506 includes:

[0339] The first constraint unit is used to construct external environmental constraint conditions based on the drivable tunnels of the target path point.

[0340] The second constraint unit is used to construct dynamic constraints for the system based on the coordinates of the rear axle center point, the front wheel deflection angle, the rate of change of the front wheel deflection angle, and the front and rear wheelbase.

[0341] The third constraint element is used to establish the first boundary value constraint conditions based on the coordinates of the starting point and the ending point.

[0342] The fourth constraint unit is used to construct the second boundary value constraint condition based on the boundary information of the map boundary and the front wheel deflection angle information of the vehicle.

[0343] The map boundary information includes the minimum and maximum values ​​of the horizontal and vertical coordinates of the map boundary, and the front wheel deflection angle information includes the upper and lower bounds of the front wheel deflection angle and the upper and lower bounds of the front wheel deflection angle transformation rate.

[0344] The cost function construction unit is used to construct the cost function based on the initial path length, the front wheel yaw angle adjustment coefficient, and the front wheel yaw angle change rate adjustment coefficient.

[0345] The fourth determining unit is used to select the path corresponding to the solution with the minimum cost function as the final path, provided that the external environment constraints, system dynamic constraints, first boundary value constraints, and second boundary value constraints are satisfied.

[0346] For details on the specific working process of the units disclosed in the above embodiments of this application, please refer to the corresponding method embodiments, such as... Figure 13 As shown, it will not be elaborated further here.

[0347] As can be seen from the above scheme, this application provides a path planning device. After the initial planning unit 1501 performs preliminary path planning in the Frenet coordinate system of the reference path to obtain the initial path, the first generation unit 1502 generates a collision envelope for the vehicle based on the type of obstacle for each obstacle in the initial path. If the first judgment unit 1503 determines that the collision envelope of the vehicle overlaps with an obstacle on the initial path, the first determination unit 1504 takes the current path point as the target path point. Then, the construction unit 1505 constructs a drivable tunnel for the target path point, obtaining the drivable tunnel for the target path point. Finally, the optimization unit 1506 optimizes the initial path based on the drivable tunnels of all target path points and preset constraints to obtain the final path. This achieves the goal of efficiently and stably planning obstacle avoidance paths when facing different obstacle types.

[0348] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0349] Another embodiment of this application provides an electronic device, such as... Figure 16 As shown, it includes:

[0350] One or more processors 1601.

[0351] Storage device 1602, on which one or more programs are stored.

[0352] When the one or more programs are executed by the one or more processors 1601, the one or more processors 1601 implement the path planning method as described in any of the above embodiments.

[0353] Another embodiment of this application provides a computer storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the path planning method as described in any of the above embodiments.

[0354] In the context of this application, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0355] It should be noted that the computer-readable medium described above in this application can be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0356] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.

[0357] Another embodiment of this application provides a computer program product, which, when executed, is used to perform any of the path planning methods described above.

[0358] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processing device, it performs the functions defined in the methods of the embodiments of this application.

[0359] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

[0360] While several specific implementation details are included in the foregoing discussion, these should not be construed as limiting the scope of this application. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0361] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described application concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions claimed in this application.

Claims

1. A path planning method characterized by, The method comprises the following steps: performing preliminary path planning in the Frenet coordinate system of the reference path to obtain an initial path; for each obstacle in the initial path, generating a collision envelope of the vehicle according to the type of the obstacle; judging whether the collision envelope of the vehicle overlaps with the obstacle on the initial path; if it is judged that the collision envelope of the vehicle overlaps with the obstacle on the initial path, taking the current path point as a target path point; constructing a drivable tunnel for the target path point to obtain a drivable tunnel of the target path point; optimizing the initial path according to the drivable tunnels of all the target path points and preset constraint conditions to obtain a final path.

2. The path planning method of claim 1, wherein, The method of performing preliminary path planning in the Frenet coordinate system of the reference path to obtain an initial path comprises the following steps: equally dividing a longitudinal axis of the Frenet coordinate system of the reference path into N sampling layers; wherein N is a positive integer; for each sampling layer, sampling M sampling points on the sampling layer at equal intervals along one side boundary to another side boundary of a transverse axis of the Frenet coordinate system of the reference path; wherein M is a positive integer; calculating the cost of each sampling point of each sampling layer and recording the parent node of the sampling point; starting from the sampling point with the minimum cost in the last sampling layer, backtracking to the starting point according to the parent node of the sampling point to obtain an initial path.

3. The path planning method of claim 2, wherein, The method of calculating the cost of each sampling point of each sampling layer comprises the following steps: determining the cost of the sampling point according to the transverse lateral displacement cost, the change cost of the transverse offset, the change rate cost of the transverse offset and the collision cost.

4. The path planning method of claim 2, wherein, If there is a cross-lane obstacle in the reference path, the method of sampling M sampling points on the sampling layer at equal intervals along one side boundary to another side boundary of a transverse axis of the Frenet coordinate system of the reference path further comprises the following steps: determining the nearest sampling point of the cross-lane obstacle on the Frenet coordinate system of the reference path according to the coordinates of the cross-lane obstacle in the Frenet coordinate system of the reference path; replacing the transverse coordinate of the nearest sampling point with the transverse coordinate of the cross-lane obstacle in the Frenet coordinate system of the reference path.

5. The path planning method of claim 4, wherein, The method of determining the nearest sampling point of the cross-lane obstacle on the Frenet coordinate system of the reference path according to the coordinates of the cross-lane obstacle in the Frenet coordinate system of the reference path comprises the following steps: determining the nearest sampling layer of the cross-lane obstacle according to the longitudinal coordinate of the cross-lane obstacle in the Frenet coordinate system of the reference path; for each sampling point in the nearest sampling layer, calculating the distance between the sampling point and the cross-lane obstacle; taking the sampling point closest to the cross-lane obstacle as the nearest sampling point of the cross-lane obstacle on the Frenet coordinate system of the reference path.

6. The path planning method of claim 2, wherein, The one side boundary and the other side boundary of the transverse axis of the Frenet coordinate system of the reference path leave a sampling threshold.

7. The path planning method of claim 1, wherein, The generating, for each of the obstacles in the initial path, a collision envelope of the vehicle according to a type of the obstacle comprises: For each of the obstacles in the initial path, if the type of the obstacle is a crossable obstacle, generating four tire envelopes; If the type of the obstacle is a detour obstacle, generating a whole vehicle envelope.

8. The path planning method of claim 1, wherein, If the type of the obstacle of the target path point is a detour obstacle, the constructing a drivable tunnel for the target path point comprises: At the target path point, generating a first disc and a second disc according to size information of the vehicle and coordinate information of a center of a rear axle of the vehicle; wherein the first disc and the second disc have the same radius; Inflating the obstacle according to the radius of the first disc and the second disc to obtain an inflated obstacle; Generating a first boundary of the drivable tunnel according to a center of the first disc, the inflated obstacle, and a preset step length; Generating a second boundary of the drivable tunnel according to a center of the second disc, the inflated obstacle, and the preset step length; Generating the drivable tunnel of the target path point according to the first boundary of the drivable tunnel and the second boundary of the drivable tunnel.

9. The path planning method of claim 8, wherein, The generating the first boundary of the drivable tunnel according to the center of the first disc, the inflated obstacle, and the preset step length comprises: Taking the center of the first disc as an initial point, sequentially expanding and constructing a rectangle in a preset step length according to a sequence of four target directions; Judging whether the rectangle overlaps with the inflated obstacle; If it is judged that the rectangle does not overlap with the inflated obstacle, combining all the currently obtained rectangles, and expanding and constructing a rectangle in a preset step length according to a sequence of target directions to a next target direction; If it is judged that the rectangle overlaps with the inflated obstacle, removing the target direction from the sequence of target directions, and continuing to expand and construct a rectangle in a preset step length according to a sequence of target directions to a next target direction; When all the rectangles constructed by the target directions all overlap with the inflated obstacle, generating the first boundary of the drivable tunnel according to all the currently obtained rectangles which do not overlap with the inflated obstacle.

10. The path planning method of claim 8, wherein, The generating the second boundary of the drivable tunnel according to the center of the second disc, the inflated obstacle, and the preset step length comprises: Taking the center of the second disc as an initial point, sequentially expanding and constructing a rectangle in a preset step length according to a sequence of four target directions; Judging whether the rectangle overlaps with the inflated obstacle; If it is judged that the rectangle does not overlap with the inflated obstacle, combining all the currently obtained rectangles, and expanding and constructing a rectangle in a preset step length according to a sequence of target directions to a next target direction; If it is judged that the rectangle overlaps with the inflated obstacle, removing the target direction from the sequence of target directions, and continuing to expand and construct a rectangle in a preset step length according to a sequence of target directions to a next target direction; When all the rectangles constructed according to the target directions overlap with the inflated obstacle, a second boundary of the drivable tunnel is generated according to all the rectangles that do not overlap with the inflated obstacle.

11. The path planning method of claim 1, wherein, If the type of the obstacle of the target path point is the crossable obstacle, the constructing of the drivable tunnel for the target path point includes: taking the target path point as a first undetermined point; moving the first undetermined point by a preset step length in the target direction to obtain a second undetermined path point in the target direction; constructing four tire envelopes corresponding to the second undetermined path point based on the second undetermined path point; judging whether the four tire envelopes corresponding to the second undetermined path point overlap with the obstacle; if it is judged that the four tire envelopes corresponding to the second undetermined path point do not overlap with the obstacle, taking the second undetermined path point as a new first undetermined path point and returning to execute the step of moving the first undetermined point by a preset step length in the target direction to obtain a second undetermined path point in the target direction; if it is judged that the four tire envelopes corresponding to the second undetermined path point overlap with the obstacle, taking the first undetermined point as a boundary of the drivable tunnel in the target direction; generating the drivable tunnel of the target path point according to the boundaries of the drivable tunnel in all the target directions.

12. The path planning method of claim 1, wherein, The optimization of the initial path according to the drivable tunnels of all the target path points and the preset constraint conditions to obtain a final path includes: constructing an external environment limit constraint condition according to the drivable tunnel of the target path point; constructing a system dynamic constraint condition according to the coordinates of the center point of the rear axle of the vehicle, the front wheel deflection angle of the vehicle, the front wheel deflection angle change rate and the front-rear wheel axle distance; establishing a first boundary value constraint condition according to the coordinates of the starting point and the coordinates of the ending point; constructing a second boundary value constraint condition according to the boundary information of the map boundary and the front wheel deflection angle information of the vehicle, wherein the boundary information of the map includes the minimum value of the horizontal and vertical coordinates of the map boundary and the maximum value of the horizontal and vertical coordinates, and the front wheel deflection angle information includes the upper limit and the lower limit of the front wheel deflection angle and the upper limit and the lower limit of the front wheel deflection angle change rate; constructing a cost function according to the initial path length, the front wheel deflection angle adjustment coefficient and the front wheel deflection angle change rate adjustment coefficient; under the conditions of satisfying the external environment limit constraint condition, the system dynamic constraint condition, the first boundary value constraint condition and the second boundary value constraint condition, taking the path corresponding to the solution with the minimum cost function as the final path.

13. A path planning device characterized by comprising: The method includes: an initial planning unit configured to perform preliminary path planning in a Frenet coordinate system of a reference path to obtain an initial path; a first generation unit configured to generate a collision envelope of a vehicle according to the type of each obstacle in the initial path; a first judgment unit configured to judge whether the collision envelope of the vehicle overlaps with the obstacle on the initial path; a first determination unit configured to take a current path point as a target path point if the first judgment unit judges that the collision envelope of the vehicle overlaps with the obstacle on the initial path; A constructing unit is configured to construct drivable tunnels for the target path points to obtain drivable tunnels for the target path points. An optimizing unit is configured to optimize the initial path according to the drivable tunnels for all the target path points and preset constraint conditions to obtain a final path.

14. An electronic device, comprising: Comprise: one or more processors; a storage device having stored thereon one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the path planning method according to any one of claims 1 to 12.

15. A computer storage medium, comprising, a computer program is stored thereon, wherein the computer program is executed by a processor to implement the path planning method according to any one of claims 1 to 12.

Images