WO2022056770A1

WO2022056770A1 - Path planning method and path planning apparatus

Info

Publication number: WO2022056770A1
Application number: PCT/CN2020/115825
Authority: WO
Inventors: 陈可际; 刘亚林
Original assignee: 华为技术有限公司
Priority date: 2020-09-17
Filing date: 2020-09-17
Publication date: 2022-03-24
Also published as: CN113286985A

Abstract

Embodiments of the present application provide a path planning method, used in the field of intelligent driving. By means of an improved A* algorithm, a smooth path which is not constrained by a grid is planned. Specifically, the method is: determining a subsequent node search area by means of a current node in a path; determining a plurality of sampling points by means of sampling in the search area, then determining the location of the subsequent node from among the plurality of sampling points. The method provided by the present application can be applied to smart vehicles, new-energy vehicles, and self-driving vehicles. The sampling points in the present method are not limited by a grid, therefore the target path obtained by planning satisfies the dynamic constraints of the target, and in comparison with the prior art, the path is smoother.

Description

Path planning method and path planning device

technical field

The present application relates to the field of intelligent driving, and in particular, to a path planning method and a path planning device.

Background technique

Intelligent driving technology is the focus of future automotive industry upgrading. The intelligent driving system can be roughly divided into three main modules, namely perception recognition, decision planning and control execution. Smart cars perceive and identify the external environment through sensors such as radar and cameras. On the basis of multi-information fusion, the intelligent driving system performs corresponding decision-making planning, including driving task decision-making, trajectory planning, and exception handling. The result of the decision planning will be passed to the lower level control execution to realize. In most intelligent driving systems, decision planning is a core module, and path planning is an important part of decision planning.

Search-based path planning methods are mostly used for global path planning or driving navigation. Common algorithms include Dijkstra algorithm, A-Star (A*) algorithm, rapid-exploration random tree (rapid-exploration random tree) , RRT) algorithm and so on. The A* search algorithm is a widely popular heuristic search algorithm. It uses the map grid as a node and introduces a cost function based on the Dijkstra algorithm, which can quickly and efficiently search for the shortest path from the start point to the end point.

As shown in Figure 1a and Figure 1b, the traditional A* search algorithm needs to traverse the adjacent nodes around the current node, and the movement of the vehicle is restricted by the grid. In the path planning of each step, the intelligent vehicle can only move from one grid center to another grid center, and the moving distance and steering angle are relatively fixed.

SUMMARY OF THE INVENTION

The embodiment of the present application provides a path planning method, which is used to obtain a smooth path that is not restricted by a grid.

A first aspect of the embodiments of the present application provides a path planning method, including: acquiring a location of an end point of a moving target in a map; determining a search area in the map according to the location of the first node; Obtain multiple sampling points from the multiple sampling points; obtain the sampling point that minimizes the cost function from the multiple sampling points as the second node, and the second node is the next node of the first node in the target path; according to the The location of the end point, the location of the first node, and the location of the second node determine the target path.

In a possible implementation manner, the first node is the starting point of the target path.

In another possible implementation manner, the first node is an intermediate node of the target path; the method further includes: acquiring the position of the starting point of the moving target in the map, and the position of the starting point is determined by to determine the target path.

In the path planning method provided by the embodiment of the present application, firstly, based on the position of the first node, that is, the current node, the search area of the next node, that is, the second node is determined, and then the next node is determined from a plurality of sampling points in the search area. Thus, It can make the position of the next node in the target path not limited by the grid in the traditional A-star algorithm, and avoid the instantaneous large turning point of the path. The distance and heading angle between the path nodes in the path obtained according to this method can be flexibly The target path is smoother and the path quality is high, which improves the practicability of the path planning algorithm.

In a possible implementation manner of the first aspect, the determining the search area in the map according to the position of the first node specifically includes: according to the position of the first node and the moving target in the The heading angle of the location of the first node determines the search area.

The path planning method provided by the embodiment of the present application can also limit the search area to be determined based on the position of the current node and the heading angle of the moving target at the current position. Since the vehicle cannot turn around or move sideways instantaneously during the normal driving process of the vehicle, Therefore, part of the area near the position of the current node does not conform to the vehicle dynamics constraints. This method determines the search area based on the heading angle of the moving target at the current position, and the acquired next node conforms to the dynamic constraints. In addition, the calculation amount can be reduced.

In a possible implementation manner of the first aspect, the method further includes: acquiring dynamic constraint information of the moving target at the position of the first node and/or information about the periphery of the position of the first node road information; the dynamic constraint information includes at least one of the following: the speed of the moving target at the position of the first node, the heading angle change rate of the moving target at the position of the first node; the The road information includes at least one of the following: the location of obstacles around the location of the first node and the location of the road boundary; the determining the search area in the map according to the location of the first node specifically includes: according to the location of the first node. The position of the first node and the heading angle of the moving target at the position of the first node, and at least one of the following: the dynamic constraint information and the road information, determine the search area.

In the path planning method provided by the embodiment of the present application, when determining the search area of the next node, the process considers the position of the current node and the heading angle of the moving target at the current node position, and can also consider dynamic constraint information and road information, The planned search area is more in line with the dynamic constraints, and can also exclude obstacles or some areas outside the road area, reducing the amount of calculation.

In a possible implementation manner of the first aspect, the length of the search area extending in the direction of the heading angle of the moving target is positively related to the current speed of the moving target.

In the path planning method provided by the embodiment of the present application, the length of the search area extending in the direction of the heading angle of the moving target can be determined according to the current speed of the moving target. The faster the speed of the moving target, the farther the moving distance per unit time. , by increasing the length of the search area extending in the direction of the heading angle, the search area can be made to meet the actual requirements.

In a possible implementation manner of the first aspect, the search area is an area with the position of the first node as a boundary point.

In the path planning method provided by the embodiments of the present application, the search area may be a sector or a triangle, and the area can be more conveniently obtained based on the preset search area shape, reducing the computational complexity of the algorithm and improving the path planning efficiency.

In a possible implementation manner of the first aspect, a search angle of the search area is less than or equal to 180 degrees, the search angle of the search area is an angle formed by two tangents passing through the first node, and the The included angle includes the direction of the heading angle of the moving target at the position of the first node.

The path planning method provided by the embodiment of the present application further limits the shape of the search area. Considering that the vehicle cannot turn around or move sideways in an instant during normal driving, the search angle of the search area is limited so that the location near the current node can be excluded. It can reduce the amount of calculation and improve the performance of the algorithm.

In a possible implementation manner of the first aspect, the size of the search angle of the search area is positively correlated with the heading angle change rate of the position of the moving target at the first node.

In the path planning method provided by the embodiment of the present application, the size of the search angle of the search area can be determined according to the current rate of change of the heading angle of the moving target. For a situation where the rate of change of the heading angle is relatively large, such as when turning, the size of the search angle can be increased to satisfy the target. Path search requirements.

In a possible implementation manner of the first aspect, the acquiring multiple sampling points in the search area specifically includes: acquiring the multiple sampling points through random sampling in the search area; The search area obtains multiple sampling points through systematic sampling.

In the path planning method provided by the embodiment of the present application, there are many methods for selecting sampling points in the search area, such as random sampling or uniform sampling of sampling points through systematic sampling, so that the position of the next node in the path determined by this method can be freed from The limitation of the grid in the traditional A-star algorithm makes the planned target path smoother and improves the path quality.

A second aspect of the embodiments of the present application provides a path planning device, including: an acquiring unit, configured to acquire the position of the end point of a moving target in a map; a determining unit, configured to determine the location on the map according to the position of the first node The search area in a second node, where the second node is the next node of the first node in the target path; the determining unit is further configured to, according to the position of the end point, the position of the first node and the second node The location of the node determines the target path.

In a possible implementation manner of the second aspect, the search area is determined according to the position of the first node and the heading angle of the moving target at the position of the first node.

In a possible implementation manner of the second aspect, the obtaining unit is further configured to: obtain the dynamic constraint information of the position of the moving target at the first node and/or the periphery of the position of the first node The dynamic constraint information includes at least one of the following: the speed of the moving target at the position of the first node, the heading angle change rate of the moving target at the position of the first node; the The road information includes at least one of the following: the position of obstacles around the position of the first node and the position of the road boundary; the determining unit is specifically configured to: according to the position of the first node and the moving target The heading angle at the position of the first node, and at least one of the following: the dynamic constraint information, the road information, determine the search area.

In a possible implementation manner of the second aspect, the length of the search area extending in the direction of the heading angle of the moving target is positively related to the current speed of the moving target.

In a possible implementation manner of the second aspect, the search area is fan-shaped or triangular.

In a possible implementation manner of the second aspect, the search area is an area with the position of the first node as a boundary point.

In a possible implementation manner of the second aspect, a search angle of the search area is less than or equal to 180 degrees, and the search angle of the search area is an angle formed by two tangents passing through the first node, and the The included angle includes the direction of the heading angle of the moving target at the position of the first node.

In a possible implementation manner of the second aspect, the size of the search angle of the search area is positively correlated with the heading angle change rate of the position of the moving target at the first node.

In a possible implementation manner of the second aspect, the acquiring unit is specifically configured to: acquire the plurality of sampling points by random sampling in the search area; or acquire the search area by systematic sampling multiple sampling points.

A third aspect of the embodiments of the present application provides a path planning apparatus, including: one or more processors and a memory; wherein, the memory stores computer-readable instructions; the one or more processors read The computer readable instructions cause the terminal to implement the method as described in any of the above first aspect and various possible implementations.

A fourth aspect of the embodiments of the present application provides a computer program product including instructions, which, when run on a computer, enables the computer to execute the first aspect and any one of various possible implementation manners. method.

A fifth aspect of the embodiments of the present application provides a computer-readable storage medium, including instructions, when the instructions are executed on a computer, the computer is made to execute the above-mentioned first aspect and any one of various possible implementation manners. method described.

A sixth aspect of the embodiments of the present application provides a vehicle, including the path planning device described in any one of the foregoing second aspect or the third aspect and various possible implementation manners.

A seventh aspect of an embodiment of the present application provides a chip, including a processor. The processor is configured to read and execute the computer program stored in the memory to perform the method in any possible implementation manner of any of the above aspects. Optionally, the chip includes a memory, and the memory and the processor are connected to the memory through a circuit or a wire. Further optionally, the chip further includes a communication interface, and the processor is connected to the communication interface. The communication interface is used for receiving data and/or information to be processed, the processor obtains the data and/or information from the communication interface, processes the data and/or information, and outputs the processing result through the communication interface. The communication interface may be an input-output interface.

Wherein, for the technical effect brought by any one of the second aspect, the third aspect, the fourth aspect, the fifth aspect, the sixth aspect or the seventh aspect, reference may be made to the technical effect brought by the corresponding implementation manner in the first aspect The technical effect will not be repeated here.

As can be seen from the above technical solutions, the embodiments of the present application have the following advantages:

In the path planning method provided by the embodiments of the present application, the search area is determined by the current node in the path, multiple sampling points are determined by sampling in the search area, and then the position of the next node is determined from the multiple sampling points. Not limited by the grid, therefore, the planned target path can satisfy the dynamic constraints of the target, and is smoother than the path in the prior art.

Description of drawings

Fig. 1a is a schematic diagram of the traditional A-star algorithm;

Figure 1b is another schematic diagram of the traditional A-star algorithm;

FIG. 2 is an architectural diagram of an application scenario of the path planning method in the embodiment of the present application;

3a is a schematic diagram of an embodiment of a path planning method in an embodiment of the present application;

3b is an algorithm flow chart of the path planning method in the embodiment of the present application;

4a is a schematic diagram of an embodiment of a search area in an embodiment of the present application;

4b is a schematic diagram of a path planning simulation effect of the path planning method provided by the embodiment of the application;

Fig. 4c is a schematic diagram of the simulation effect of path planning obtained based on the traditional A* search algorithm;

FIG. 5a is a schematic diagram of another embodiment of the search area in the embodiment of the present application;

FIG. 5b is another schematic diagram of the path planning simulation effect of the path planning method provided by the embodiment of the application;

FIG. 6 is a schematic diagram of an embodiment of a path planning apparatus in an embodiment of the present application;

FIG. 7 is a schematic diagram of another embodiment of the path planning apparatus in the embodiment of the present application.

detailed description

For ease of understanding, some technical terms involved in the embodiments of the present application are briefly introduced below:

1. Open list: used to save all nodes that have been generated but not investigated;

2. Close list: used to record the nodes that have been visited;

3. The cost function F(n) is a function to calculate the search cost. The value of F(n) obtained indicates the importance of the current node. The smaller F(n) is, the cost to reach the end point after passing through the node. The smaller (the shorter the distance), the stronger the importance of the node. The expression of the cost function of the A-star algorithm is: F(n)=G(n)+H(n).

Among them, G(n): used to indicate the cost that has been paid from the starting node to the current node n, for example: the shortest path value from the starting point to the node n; H(n): used to indicate the cost from the current node n to the end point the price to pay. For example: the shortest path value from node n to the end point.

The expression of the cost function in the embodiment of the present application may also be expressed as: F=G+H.

Optionally, where G represents the distance from the sampling point to the previous node and the change in heading angle; H is the shortest path from the sampling point to the end point.

4. Systematic sampling: According to a certain sampling distance, samples are drawn from the population. Specifically, it is a sampling method that divides the population into balanced parts, and then selects an individual from each part according to predetermined rules to obtain the required sample.

The A* algorithm is briefly introduced below, please refer to Figure 1a-Figure 1b:

The A* search algorithm is a widely popular heuristic search algorithm. It uses the map grid as a node and introduces a cost function based on the Dijkstra algorithm, which can quickly and efficiently search for the shortest path from the start point to the end point.

The basic steps of the traditional A* search algorithm are as follows:

1. Add the starting point to the Open list, and the Close list is initialized to be empty.

2. Determine whether the Open list is empty. If so, the algorithm ends and the path planning fails; if not, proceed to the next step.

3. Take the node with the smallest cost function value from the Open list as the next exploration node P, and move it into the Close list.

4. Determine whether the node P is the target point (ie the end point). If yes, go to step 8 directly; if no, go to the next step.

5. Search for adjacent nodes Pi around node P, and perform step 6.

6. If Pi is not in the Close list, calculate the cost function value of the Pi node and add it to the Open list. If the Pi node is already in the Open list, and the cost function value recorded by the Open list is greater than the current calculated value, then update the information recorded by the Pi node in the Open list.

7. After traversing all Pi nodes, go back to step 2.

8. Search the parent node information in the Close list in reverse order to get the global path from the start point to the end point.

As shown in Figure 1a, the A* algorithm can not only search for the shortest distance from the starting point to the ending point, but also plan the shortest obstacle avoidance path when the obstacle information is known.

As shown in Figure 1b, the traditional A* algorithm builds a grid in the map, and the nodes in the planned path are all grid centers. When searching for the node at the end of the path, the current node position will be traversed to investigate the adjacent 8 nodes around the current node position (1-8 in Figure 1b), regardless of the driving state of the vehicle. For example, in Figure 1b, the vehicle is currently driving The direction is that the front of the vehicle points to node 5. During normal driving, since the vehicle cannot turn around or move sideways, nodes 1, 2, 4, 6 and 7 obviously do not meet the vehicle dynamics constraints. It wastes computing resources and has low search efficiency. In addition, since the nodes in the path are all the center points of the grid, the planned path, in the position shown in Figure 1b, restricts the vehicle to only drive straight forward to node 5, or to deflect 45 degrees to the left and forward to the node 3 Drive, or, yaw 45 degrees ahead to the right toward node 8. It is not difficult to understand that the resulting path, the trajectory machine, may have an instantaneous large turning point.

In some search-based path planning methods, the vehicle dynamics model is combined to improve the search algorithm, such as the hybrid A-star algorithm (hybrid A*), the variant of the rapidly expanding random tree (RRT) algorithm, and so on. These methods add the calculation of the dynamic model to the algorithm process, which can effectively screen out the search nodes that meet the vehicle dynamic constraints, and avoid the instantaneous large turning point in the vehicle driving path. However, in the path search algorithm combined with the vehicle dynamics model, the steps of iterative operation of the model are added, the calculation is complicated, and the search efficiency is reduced.

The path planning method provided by the embodiments of the present application can be applied to various moving targets (including motor vehicles, non-motor vehicles, boats, boats, pedestrians or robots, etc.) in various forms of paths (including highways, urban roads, and rural roads). or indoor paths, etc.), such as the route planning of unmanned ships and unmanned speedboats. Subsequent embodiments are described by taking the path planning of the vehicle on the road as an example, but those skilled in the art can extend it to the field of path planning for other targets, which is not specifically limited here.

2, the architecture diagram of the application scenario of the path planning method in the embodiment of the present application, the network elements involved in the present application are briefly introduced:

(1) Positioning and perception sensors, such as cameras, lidars, etc.;

(2) Vehicle position and attitude sensors, such as inertial measurement unit (IMU), global positioning system (global positioning system, GPS), etc.;

(3) In-vehicle gateway, T-BOX such as front loader (telematics box, T-Box) is mainly used to communicate with background system/mobile APP to realize vehicle information display and control of mobile APP;

(4) In-vehicle data computing unit, such as mobile data center (MDC), etc.;

(5) Cloud data computing units, such as cloud servers, etc.

The path planning method proposed in the present application is implemented by a path planning device, and the path planning device may be an on-board data computing unit or a cloud data computing unit, which is not specifically limited here. Optionally, when the computing resources of the on-board data computing unit in the vehicle are tight or malfunction occurs, the path planning algorithm of the present application can be switched to the cloud data computing unit for execution, so as not to affect the normal driving of the intelligent vehicle.

The path planning method provided by the embodiment of the present application is introduced below, please refer to FIG. 3a, the method includes:

3011. The path planning device acquires the position of the end point of the moving target in the map;

A map can be preset in the path planning device, or a map of a corresponding geographic location area can be obtained online according to the real-time position of the moving target. The specific type of the map is not limited, and it can be a normal map or a high-precision map. It can be understood that the current position of the moving target, the position of the end point and the target path to be planned are all located in the area covered by the map. Before performing path planning, it is also necessary to obtain the position of the end point of the moving target and the position of the first node. The target path is a path expected to be obtained by the path planning apparatus, and the first node may be any intermediate node in the target path, or the starting point of the target path. In this embodiment, the first node is taken as the starting point of the target path as an example for introduction.

3012. Determine a search area in the map according to the position of the first node;

In the path planning method of the embodiment of the present application, the position of the intermediate node of the path needs to be determined, and then the target path is determined according to the position of the starting point, the position of the intermediate node, and the position of the end point, and the method for determining the path based on multiple nodes in the path, such as the difference value The method and the like are the prior art, which is not specifically limited here. It should be noted that in addition to the known position of the starting point and the position of the ending point, the path planning device also needs to obtain at least one intermediate node in the target path. It can be understood that, in order to determine the target path, it is usually necessary to search for the starting point and the ending point. For the convenience of description, in this embodiment, according to the order in the target path, each node starting from the starting point is sequentially called the first node, the second node...the Nth node, according to the first node. The position of the node can determine the second search area of the second node, and so on, until the position of the end point in the Nth search area of the Nth node is determined, and the search can be ended. The method for determining the search area from the first node to the Nth node is similar to the method for determining the search area for the first node, and details are not repeated here. The following takes the search process of the second node as an example for introduction.

The path planning device determines the position of the next node in the path, ie, the second node, according to the position of the first node. First, the search area is determined according to the position of the first node. The following describes the determination method of the search area.

There are many methods for determining the search area. Optionally, the path planning device determines the search area according to the position (x ₀ , y ₀ ) of the first node and the heading angle (θ ₀ ) of the moving target at the position of the first node . Here, the heading angle of the moving target at the starting point may be a preset value, and the specific direction is not limited. The heading angle of the target vehicle at the position of the intermediate node can be calculated according to the position of the previous node and the position of the current node.

Optionally, in addition to the position of the first node (x ₀ , y ₀ ) and the heading angle (θ ₀ ) of the target vehicle when the target vehicle is at the position of the first node, the path planning device also bases on the power of the target vehicle at the current first node The search area is determined by the learning constraint information, and the dynamic constraint information includes the speed, the rate of change of the heading angle, etc., which reflects the motion state of the target vehicle. Optionally, the speed or heading angle change rate of the target vehicle may be a preset value.

Optionally, the path planning device also acquires road environment information, such as the location of obstacles or the location of road boundaries, which can be used to determine the search area.

The shape of the search area is not limited, optionally, the search area is an area with the position of the first node as the boundary point, the search angle of the search area is less than or equal to 180 degrees, and the search angle of the search area is over The included angle formed by the two tangents of the first node and including the heading angle direction. Optionally, the search area is a preset shape, such as a fan shape or a triangle shape.

The size of the search area may be a preset size, or determined according to the motion state of the target vehicle, which is not limited here. Optionally, the length of the search area extending in the direction of the heading angle is positively correlated with the current speed of the target vehicle. Optionally, the size of the search angle of the search area is positively correlated with the current rate of change of the heading angle of the target vehicle, and the search angle of the search area is the heading angle formed by the two tangents passing through the first node. angle of direction.

3013. Determine multiple sampling points in the search area;

The path planning device determines a plurality of sampling points in the search area. The number of sampling points is at least two, and the specific value is related to the search area and search accuracy, and the specific number is not limited here.

There are various methods for selecting sampling points. Optionally, the path planning device determines the plurality of sampling points by random sampling in the search area; or, the path planning device determines the plurality of sampling points in the search area by systematic sampling. Sampling point. The specific method for selecting sampling points is not limited here.

3014. Determine, from the plurality of sampling points, the sampling point that minimizes the cost function as the first node;

The path planning apparatus traverses the plurality of sampling points Pi, and determines that the center point that minimizes the cost function is the first intermediate node, and the specific calculation formula of the cost function is not limited in this embodiment of the present application. Optionally, the expression of the cost function in the embodiment of the present application may also be expressed as: F=G+H. Among them, G represents the distance from the sampling point to the previous node and the change of the heading angle; H is the shortest path from the sampling point to the end point.

3015. Determine the target path according to the position of the first node, the position of the second node, and the position of the end point;

The path planning device determines the target path according to the position of the first node, the position of the second node and the position of the end point. It can be understood that the number of the first nodes is not limited and can be one or more.

The path planning method provided by the embodiment of the present application specifically includes determining a search area for the next node by using a current node in the path, determining multiple sampling points by sampling in the search area, and then determining the position of the next node from the multiple sampling points , since the sampling points in this method are not limited by the grid, the planned target path can satisfy the dynamic constraints of the target, and is smoother than the path in the prior art.

It should be noted that, the path planning method provided by the embodiment of the present application is improved on the basis of the traditional A-star algorithm. As shown in FIG. 3b, the algorithm flow steps of the path planning method in the embodiment of the present application are as follows:

3021. Add the starting point to the Open list.

3022. Determine whether the Open list is empty, if so, go to step 3023; if not, go to step 3024

3023. If the Open list is empty, the path planning fails and the program exits;

3024. If the Open list is not empty, find the node with the smallest cost function value from the Open list as the next exploration node P.

3025. Determine whether the node P is the end point, if yes, go to step 3026; if not, go to step 3027.

3026. If the node P is the end point, search the parent node information in the Close list in reverse order to obtain the global path from the start point to the end point.

3027. If the node P is not the end point, determine the search area of the next node, and determine multiple sampling points Pi within the area to traverse.

It should be noted that, if the search area of the next node covers the location of the end point, the end point can be directly used as the only sampling point, that is, the next node is determined.

3028. If Pi is in the Close list, skip the sampling point; calculate the cost function value of Pi, and add the sampling point information to the Open list.

Optionally, in the algorithm flow of the embodiment of the present application, auxiliary calculation may also be performed based on grids, and the information contained in the Open list and the Close list include: the horizontal and vertical coordinates of the grid occupied by the sampling points (x _g , y _{g )} ), the horizontal and vertical coordinates of the sampling point itself (x _s , y _s ) and the heading angle θ _s , the horizontal and vertical coordinates of the grid occupied by the parent node (x _gp , y _gp ), the horizontal and vertical coordinates of the parent node itself (x _sp , y _sp ) and the heading angle θ _sp , the cost function value F. In step 3028, if the grid occupied by the sampling point to be investigated is already in the Open list, and the cost function value recorded in the Open list is greater than the current value, then update the corresponding sampling point information and the cost function value.

Steps 3022 to 3028 are repeated until the global path from the start point to the end point is obtained.

The shape of the search area in the path planning method according to the embodiment of the present application may be of various types, and the following description is given by taking a sector shape and a triangle shape as examples.

Please refer to Fig. 4a, Fig. 4a shows a sector search area in an embodiment of the present application;

In this embodiment, a sector is used as the search area, the center of the sector coincides with the current coordinates of the vehicle, and the radius L and the center angle θ _max are related to the vehicle pose, vehicle dynamics constraints, and road environment information.

1) Determine the search area;

Exemplarily, if only the vehicle speed is considered, the position of the fan-shaped area can be calculated according to the following formula.

When the vehicle dynamics constraints need to be strengthened, for example, when the vehicle speed is greater than or equal to a preset threshold, L can be increased and θ _max can be decreased. When the road environment is complex, for example, when there are many obstacles and the road curvature is large, L can be decreased and θ _max can be increased.

L=L ₀ +α*V

θ _max = θ _max0 -V/β*π

Among them, V is the vehicle speed (m/s); L ₀ is a preset value, and the value range is, for example, 3 to 5 (meters); θ _max0 can be π; α is a preset coefficient, and the value range is [0, 1] , such as 0.5; β is also a preset coefficient, and the specific value is not limited, such as 50.

2) Sampling in the search area to obtain multiple sampling points;

In this embodiment, quantitative sampling is performed in the sector search area, that is, the radius and the central angle of the sector are quantitatively divided. On the basis of knowing the current node coordinates (x ₀ , y ₀ ) and the heading angle θ ₀ , the following formulas can be used to obtain the coordinates (x _s , y _s ) of the sampling point and the heading angle θ _s .

θ _s = θ ₀ +Δθ

x _s = x ₀ +l*cosθ _s

y _s =y ₀ +l*sinθ _s

Wherein, Δθ=-θ _max /2,...,-θ _max /N ₁ , 0, θ _max /N ₁ ,..., θ _max /2;

l=L/N ₂ , 2L/N ₂ , 3L/N ₂ , ...L; l represents the distance from the current node to the sampling point.

N ₁ is the number of divisions of the central angle of the sector, and N ₂ is the number of divisions of the radius of the sector.

3) Calculate the cost function of multiple sampling points;

The cost function value F in this embodiment is calculated by the following formula.

F=G+H

G=k ₁ *l+k ₂ *|Δθ|

Wherein, k ₁ and k ₂ are weight coefficients, both of which are non-negative and usually k ₁ <k ₂ ; (x _d , y _d ) are the coordinates of the end point.

Fig. 4b shows the simulation effect of path planning obtained according to the path planning method provided in the embodiment of the present application, and Fig. 4c shows the simulation effect of path planning obtained based on the traditional A* search algorithm;

The path planning method enables the smart car to move smoothly from the starting point to the ending point, neither exceeding the road boundary nor coming into contact with obstacles. In addition, the heading angle of the path changes more flexibly, and at the same time, it is more in line with the vehicle dynamics constraints, and there are no 90° bends like many 90° bends in the results of the traditional A* search algorithm. The path nodes in the result of Example 1 of this solution are obtained by sampling in the fan-shaped search area, and are no longer limited by the map grid, so the number of path nodes in the traditional A* search is much less.

Please refer to Fig. 5a, Fig. 5a shows the triangular search area in the embodiment of the present application;

In the embodiment of this application, the search area is an isosceles triangle as an example for illustration. The vertex corresponding to the vertex angle of the _isosceles triangle coincides with the current coordinates of the vehicle. constraints and road environment information. When the vehicle dynamic constraints need to be strengthened, such as when the vehicle speed is high, L can be increased and θ _max can be decreased. When the road environment is complex, for example, when there are many obstacles and the road curvature is large, L can be decreased and θ _max can be increased.

As shown in Fig. 5a, this embodiment performs random sampling in the triangular search area, and assumes that the sampling points are uniformly distributed in the triangular search area. On the basis of knowing the current node coordinates (x ₀ , y ₀ ) and the heading angle θ ₀ , the following formulas can be used to obtain the coordinates (x _s , y _s ) of the sampling point and the heading angle θ _s .

θ _s = θ ₀ +acrtan(d/l)

x _s = x ₀ +l*cosθ _s

y _s =y ₀ +l*sinθ _s

Among them, l obeys a uniform distribution on [0,L]; d obeys a uniform distribution on [-l*tan(θ _max /2),l*tan(θ _max /2)].

Then, according to the sampling point coordinates (x _s , y _s ), the corresponding grid coordinates (x _g , y _g ) can be obtained in the map.

F=G+H

G=k ₁ *l+k ₂ *|d|

Figure 5b shows the path planning simulation effect obtained according to the path planning method provided by the embodiment of the present application. Compared with the traditional A* search algorithm, the path obtained according to the path planning method of this embodiment is smoother and more suitable for the vehicle kinetic laws. Because the path nodes are obtained by random sampling in the triangle search area, the path nodes are no longer restricted by the map grid, which improves the randomness of the path search.

The path planning method in the embodiments of the present application has been described above, and a path planning apparatus for implementing the method is introduced below.

Please refer to FIG. 6 , which is a schematic diagram of an embodiment of a path planning apparatus in an embodiment of the present application.

Only one or more of the various modules in FIG. 6 may be implemented in software, hardware, firmware, or a combination thereof. The software or firmware includes, but is not limited to, computer program instructions or code, and can be executed by a hardware processor. The hardware includes, but is not limited to, various types of integrated circuits, such as a central processing unit (CPU), a digital signal processor (DSP), a field programmable gate array (FPGA), or an application specific integrated circuit (ASIC).

The path planning device includes:

Obtaining unit 601, for obtaining the position of the end point of the moving target in the map;

a determining unit 602, configured to determine a search area in the map according to the position of the first node;

The obtaining unit 601 is further configured to: obtain a plurality of sampling points in the search area;

The obtaining unit 601 is further configured to: obtain a sampling point from the plurality of sampling points that minimizes the cost function as a second node, and the second node is the next node of the first node in the target path;

The determining unit 602 is further configured to determine the target path according to the position of the end point, the position of the first node and the position of the second node.

Optionally, the determining unit 602 is specifically configured to: determine the search area according to the position of the first node and the heading angle of the moving target at the position of the first node.

Optionally, the obtaining unit 601 is further configured to: obtain dynamic constraint information of the moving target at the position of the first node and/or road information around the position of the first node; the dynamic The learning constraint information includes at least one of the following: the speed of the moving target at the position of the first node, the heading angle change rate of the moving target at the position of the first node; the road information includes at least one of the following Item: the position of the obstacle around the position of the first node, the position of the road boundary; the determining unit 602 is specifically configured to: according to the position of the first node and the moving target at the first node The heading angle of the location, and at least one of the following: the dynamic constraint information, the road information, to determine the search area.

Optionally, the length of the search area extending in the direction of the heading angle of the moving target is positively correlated with the current speed of the moving target.

Optionally, the search area is fan-shaped or triangular.

Optionally, the search area is an area with the position of the first node as a boundary point.

Optionally, the search angle of the search area is less than or equal to 180 degrees, the search angle of the search area is an angle formed by two tangents passing through the first node, and the angle includes the moving target at The direction of the heading angle of the position of the first node.

Optionally, the size of the search angle of the search area is positively correlated with the rate of change of the heading angle of the moving target at the position of the first node.

Optionally, the acquiring unit 601 is specifically configured to: acquire the multiple sampling points in the search area by random sampling; or acquire multiple sampling points in the search area by systematic sampling.

Please refer to FIG. 7 , which is a schematic diagram of another embodiment of the path planning apparatus in the embodiment of the present application;

The path planning apparatus provided in this embodiment may be an electronic device such as a server or a terminal, and the specific device form thereof is not limited in this embodiment of the present application.

The path planning apparatus 700 may vary greatly due to different configurations or performances, and may include one or more processors 701 and a memory 702 in which programs or data are stored.

Wherein, the memory 702 may be volatile storage or non-volatile storage. Optionally, the processor 701 is one or more central processing units (CPU, Central Processing Unit, the CPU may be a single-core CPU, or a multi-core CPU. The processor 701 may communicate with the memory 702, and the path planning device 700 executes a series of instructions in memory 702.

The path planning apparatus 700 also includes one or more wired or wireless network interfaces 703, such as Ethernet interfaces.

Optionally, although not shown in FIG. 7 , the path planning apparatus 700 may also include one or more power supplies; one or more input and output interfaces, and the input and output interfaces may be used to connect a display, a mouse, a keyboard, a touch screen device or a transmission device. Sensor devices, etc., the input and output interfaces are optional components, which may or may not exist, and are not limited here.

For the process performed by the processor 701 in the path planning apparatus 700 in this embodiment, reference may be made to the method process described in the foregoing method embodiments, and details are not repeated here.

Claims

A path planning method, comprising:

Get the location of the end point of the moving target in the map;

determining a search area in the map according to the position of the first node;

acquiring a plurality of sampling points in the search area;

The sampling point obtained from the plurality of sampling points that minimizes the cost function is the second node, and the second node is the next node of the first node in the target path;

The target path is determined according to the position of the end point, the position of the first node and the position of the second node.
The method of claim 1, wherein:

The determining of the search area in the map according to the position of the first node specifically includes:

The search area is determined according to the position of the first node and the heading angle of the moving target at the position of the first node.
The method according to claim 2, wherein the method further comprises:

acquiring dynamic constraint information when the moving target is at the position of the first node and/or road information around the position of the first node;

The dynamic constraint information includes at least one of the following: the speed of the moving target at the position of the first node, and the heading angle change rate of the moving target at the position of the first node;

The road information includes at least one of the following: the location of obstacles around the location of the first node, and the location of a road boundary;

The determining of the search area in the map according to the position of the first node specifically includes:

The search area is determined according to the position of the first node and the heading angle of the moving target at the position of the first node, and at least one of the following: the dynamic constraint information and the road information.
The method according to any one of claims 1 to 3, wherein the length of the search area extending in the direction of the heading angle of the moving target is positively correlated with the current speed of the moving target.
The method according to any one of claims 1 to 4, wherein the search area is a sector or a triangle.
The method according to any one of claims 1 to 5, wherein the search area is an area with the position of the first node as a boundary point.
The method according to claim 6, wherein the search angle of the search area is less than or equal to 180 degrees, and the search angle of the search area is an angle formed by two tangents passing through the first node, so The included angle includes the direction of the heading angle of the moving target at the position of the first node.
The method according to claim 7, wherein the size of the search angle of the search area is positively correlated with the heading angle change rate of the position of the moving target at the first node.
The method according to any one of claims 1 to 8, wherein,

The acquiring a plurality of sampling points in the search area specifically includes:

acquiring the plurality of sampling points by random sampling in the search area; or,

A plurality of sampling points are acquired for the search area by means of systematic sampling.
A path planning device, comprising:

an acquisition unit, used to acquire the position of the end point of the moving target in the map;

a determining unit, configured to determine a search area in the map according to the position of the first node;

The obtaining unit is further configured to: obtain a plurality of sampling points in the search area;

The obtaining unit is further configured to: obtain a sampling point from the plurality of sampling points that minimizes the cost function as a second node, and the second node is the next node of the first node in the target path;

The determining unit is further configured to determine the target path according to the position of the end point, the position of the first node and the position of the second node.
The device according to claim 10, wherein the determining unit is specifically configured to:

The search area is determined according to the position of the first node and the heading angle of the moving target at the position of the first node.
The device according to claim 11, wherein the acquiring unit is further configured to:

acquiring dynamic constraint information of the moving target at the position of the first node and/or road information around the position of the first node;

The dynamic constraint information includes at least one of the following: the speed of the moving target at the position of the first node, and the heading angle change rate of the moving target at the position of the first node;

The road information includes at least one of the following: the location of obstacles around the location of the first node, and the location of a road boundary;

The determining unit is specifically configured to: according to the position of the first node and the heading angle of the moving target at the position of the first node, and at least one of the following: the dynamic constraint information, the road information to determine the search area.
The device according to any one of claims 10 to 12, wherein the length of the search area extending in the direction of the heading angle of the moving target is positively correlated with the current speed of the moving target.
The apparatus according to any one of claims 10 to 13, wherein the search area is a sector or a triangle.
The apparatus according to any one of claims 10 to 14, wherein the search area is an area with the position of the first node as a boundary point.
The device according to claim 15, wherein the search angle of the search area is less than or equal to 180 degrees, and the search angle of the search area is an angle formed by two tangents passing through the first node, so The included angle includes the direction of the heading angle of the moving target at the position of the first node.
The device according to claim 16, wherein the size of the search angle of the search area is positively correlated with the heading angle change rate of the position of the moving target at the first node.
The device according to any one of claims 10 to 17, characterized in that:

The obtaining unit is specifically used for:

acquiring the plurality of sampling points by random sampling in the search area; or,

A plurality of sampling points are acquired for the search area by means of systematic sampling.
A path planning device, comprising: one or more processors and a memory; wherein,

computer-readable instructions are stored in the memory;

The computer readable instructions are read by the one or more processors to cause the terminal to implement the method of any of claims 1 to 9.
A computer program product, characterized by comprising computer-readable instructions, which, when executed on a computer, cause the computer to perform the method according to any one of claims 1 to 9.
A computer-readable storage medium, characterized by comprising computer-readable instructions, which, when the computer-readable instructions are executed on a computer, cause the computer to execute the method according to any one of claims 1 to 9 .
A vehicle, characterized by comprising the route planning device according to any one of claims 10 to 18 , or the route planning device according to claim 19 .