Key point sampling method and device for spatial-temporal joint planning initial solution

By adopting the key point sampling method of space-time joint planning in the planning of autonomous driving trajectory, the problem of large amount of initial solution calculation in three-dimensional space is solved, the search speed is improved and the system delay is reduced, and more efficient path planning is achieved.

CN118938890APending Publication Date: 2024-11-12HEFEI ZHIXINGZHE TECH CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202310527062.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-05-10
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing autonomous driving trajectory planning has a large amount of initial solution calculation in three-dimensional space, resulting in slow search speed and delayed system response.

Method used

The key point sampling method of joint space-time planning is adopted, and the space-time corridor is established, obstacles are identified, and the collection point range is expanded according to the preset safety distance, and samples are taken within the collection point range, reducing the number of sampling points and reducing the number of condition judgments.

Benefits of technology

The search speed of three-dimensional spatial path planning is improved, the system communication delay is reduced, and the initial solution with practical guidance is generated, supporting subsequent convex optimization to generate smooth trajectories.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118938890A_ABST
    Figure CN118938890A_ABST
Patent Text Reader

Abstract

The invention relates to a key point sampling method and device for a spatio-temporal joint planning initial solution. The method comprises the following steps: establishing a space-time corridor in a vehicle driving process; recognizing an obstacle to obtain first position information of the obstacle in the space-time corridor; expanding the first position information according to a preset safety distance to obtain second position information after the first position information is expanded, and generating an acquisition point range based on the second position information; in the range of the acquisition points, sampling the periphery of the obstacle to obtain target key points; and when the range of the collection points is exceeded, sampling is carried out on the road center line to obtain a target key point. The scheme provided by the invention can accelerate the process of searching the initial solution in three-dimensional space planning, improves the searching speed of path planning, and reduces the possibility of system communication frequency reduction.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of path planning for autonomous driving, and in particular to a key point sampling method and device for an initial solution of spatiotemporal joint planning. Background Art

[0002] With the rapid development of science and technology, autonomous driving technology has gradually penetrated into daily life. The task of autonomous driving trajectory planning is mainly to use map data, perception data, positioning data, vehicle status and other information to calculate a collision-free executable trajectory to ensure that the autonomous driving vehicle drives safely from the starting point to the destination and is as efficient as possible. The existing autonomous driving trajectory planning framework mainly adopts decoupled planning of the Frenet coordinate system and trajectory planning directly in three-dimensional space and time. The former uses s as the direction of the road centerline and l as the direction perpendicular to the road centerline. The latter plans the longitudinal S, lateral l, and time t at the same time, and takes the path and speed as variables of the problem at the same time to obtain an optimal combination. The latter directly considers maneuvers in space and time and directly handles planning problems in three-dimensional space, so the result is generally globally optimal. However, the speed of the three-dimensional trajectory planner is slower than that of the traditional method, but it can better balance the relationship between calculation speed and solution quality.

[0003] In the process of obtaining the initial solution of the spatiotemporal joint planning in three-dimensional space, a dynamic search will be performed based on the sampling points. It is necessary to consider the kinematic constraints, curvature constraints, collision detection, traffic regulations constraints and other conditions between the sampling points to finally obtain the optimal solution. Since the search is performed in three-dimensional space, if uniform grid sampling is used on all axes, there will be thousands or even tens of thousands of condition checks during the search process, which consumes a lot of time. The initial solution will be used for convex optimization later, which will lead to a reduction in the frequency of the overall system communication and a delay in the response of the autonomous driving system.

[0004] Therefore, in the prior art, how to quickly search for an effective initial solution in three-dimensional space planning is an urgent problem to be solved. Summary of the invention

[0005] In order to solve or partially solve the problem of large amount of calculation of initial solution in three-dimensional space planning in related technologies, the present application provides a method for accelerating the search for initial solution in three-dimensional space planning, thereby improving the search speed of path planning in three-dimensional space and reducing the possibility of system communication frequency reduction.

[0006] The first aspect of the present application provides a key point sampling method for an initial solution of spatiotemporal joint planning, comprising:

[0007] Establish a space-time corridor during vehicle driving, which includes an s-axis generated in the direction of the road centerline, an l-axis perpendicular to the s-axis on the map plane, and a t-axis used to represent time. The s-axis, l-axis and t-axis are perpendicular to each other.

[0008] Identify obstacles and obtain the first position information of obstacles in the space-time corridor;

[0009] Expanding the first location information according to a preset safety distance to obtain second location information after the first location information is expanded, and generating a collection point range based on the second location information;

[0010] Within the range of the collection points, samples are taken around the obstacles to obtain the target key points; when the collection points are exceeded, samples are taken on the center line of the road to obtain the target key points.

[0011] In one embodiment, identifying obstacles includes: uniformly sampling in the s-axis direction of the space-time corridor, the l-axis direction of the space-time corridor, and the t-axis direction of the space-time corridor to identify obstacles.

[0012] In one embodiment, the first position information is expanded according to a preset safety distance to obtain second position information after the first position information is expanded, including: expanding the first position information according to the preset safety distance; extending the first position information along the l-axis and in the direction away from the l-axis by a first distance respectively to obtain a first series of coordinates and a second series of coordinates, the first distance being the half-width length of the vehicle plus the preset safety distance; extending the first position information in the s-axis direction by a second distance to obtain a third series of coordinates, the second distance being the rear wheelbase length of the vehicle plus the preset safety distance; extending the first position information in the direction away from the s-axis by a third distance to obtain a fourth series of coordinates, the third distance being the front wheelbase length of the vehicle plus the safety distance.

[0013] In one embodiment, generating a collection point range based on the second position information includes: obtaining a set of sampling variable distances, selecting the same sampling variable distance from the set, expanding the third series of coordinates and the fourth series of coordinates on the S axis, and generating a series of coordinates for the collection point range on the s axis; identifying a range of adjacent lane lines of the first series of coordinates and the second series of coordinates in a direction away from obstacles, eliminating the coordinate range between the first series of coordinates and the second series of coordinates within the adjacent lane line range, and generating a series of coordinates for the collection point range on the l axis.

[0014] In one embodiment, before generating a collection point range based on the second position information, it includes: identifying the second position information after expansion of multiple obstacles, and when the second position information after expansion of multiple obstacles overlaps, merging and calculating the second position information after expansion of the multiple obstacles; and generating a collection point range based on the merged second position information.

[0015] In one embodiment, within the range of collection points, sampling is performed around obstacles to obtain target key points, including: identifying the road centerline of the lane line, sampling the road centerline and the first series of coordinates and the second series of coordinates within the range of collection points; if the range of collection points exceeds the road boundary, then the sampling is abandoned within the range of collection points exceeding the road boundary.

[0016] In one embodiment, when the collection point range is exceeded, sampling is performed on the road centerline to obtain target key points, including: identifying the road centerline that exceeds the collection point range, and sampling equidistantly on the road centerline to obtain target key points.

[0017] The second aspect of the present application provides a key point sampling device for an initial solution of spatiotemporal joint planning, comprising:

[0018] A construction module is used to establish a space-time corridor during vehicle driving, wherein the space-time corridor includes an s-axis generated in the direction of the road centerline, an l-axis perpendicular to the s-axis on a map plane, and a t-axis used to represent time, wherein the s-axis, the l-axis and the t-axis are perpendicular to each other;

[0019] An identification module, used to identify an obstacle and obtain first position information of the obstacle in the space-time corridor;

[0020] An expansion module, configured to expand the first location information according to a preset safety distance to obtain second location information after the first location information is expanded, and generate a collection point range based on the second location information;

[0021] The sampling module is used to sample around the obstacle within the range of the collection points to obtain the target key points; when the range of the collection points is exceeded, the sampling module is used to sample on the center line of the road to obtain the target key points.

[0022] A third aspect of the present application provides an electronic device, including:

[0023] Processor; and

[0024] The memory stores executable codes, and when the executable codes are executed by the processor, the processor executes the above method.

[0025] A fourth aspect of the present application provides a computer-readable storage medium having executable code stored thereon. When the executable code is executed by a processor of an electronic device, the processor executes the above method.

[0026] The fifth aspect of the present application provides a mobile tool, including the electronic device provided by the third aspect above.

[0027] The technical solution provided by this application may have the following beneficial effects:

[0028] First, establish a space-time corridor during vehicle driving, identify obstacles through the space-time corridor, and obtain the first position information of the obstacle in the space-time corridor; expand the first position information according to the preset safety distance to obtain the second position information after the expansion of the first position information, and generate the collection point range based on the second position information. By setting the collection point range for key points, the number of sampling points can be reduced compared to the traditional grid sampling method, thereby reducing the number of process condition judgments in general; sampling around obstacles can reduce the number of meaningless detections caused by collisions.

[0029] The second aspect of the present application is to sample around obstacles within the range of the collection points to obtain the target key points; when the collection points are beyond the range, the target key points are sampled on the center line of the road. Through the preset spatial topological relationship, the key points are adopted within the range of different collection points. Even if the sampling points are rougher, the topological relationship of the lane lines is retained. The sampling points of this application can also make the searched initial solution have practical guiding significance. Subsequently, the passable corridor and convex optimization can be used to generate a smooth trajectory that satisfies curvature and kinematics.

[0030] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The above and other objects, features and advantages of the present application will become more apparent by describing in more detail exemplary embodiments of the present application in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments of the present application.

[0032] Figure 1 It is a flow chart of a key point sampling method for an initial solution of spatiotemporal joint planning shown in an embodiment of the present application;

[0033] Figure 2 is a three-dimensional spatial schematic diagram of a static obstacle in a space-time corridor shown in an embodiment of the present application;

[0034] Figure 3 is a three-dimensional spatial schematic diagram of a dynamic obstacle in a space-time corridor shown in an embodiment of the present application;

[0035] Figure 4 is a schematic diagram of first position information of an obstacle shown in an embodiment of the present application;

[0036] Figure 5 is a schematic diagram of second position information of an obstacle shown in an embodiment of the present application;

[0037] Figure 6 is a schematic diagram of the range of collection points shown in the embodiment of the present application;

[0038] Figure 7 is a schematic diagram of a collection point of an obstacle shown in an embodiment of the present application;

[0039] Figure 8 is a schematic diagram of collection points of multiple obstacles shown in an embodiment of the present application;

[0040] Fig. 9 This is a diagram showing the effect of key point sampling for the initial solution of spatiotemporal joint planning shown in the embodiment of the present application.

[0041] Fig.10 It is a structural schematic diagram of a key point sampling device for an initial solution of spatiotemporal joint planning shown in an embodiment of the present application;

[0042] Fig.11 It is a schematic diagram of the structure of an electronic device shown in an embodiment of the present application. DETAILED DESCRIPTION

[0043] The embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to make the present application more thorough and complete, and to fully convey the scope of the present application to those skilled in the art.

[0044] The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The singular forms of "a", "a", "an" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in this article refers to and includes any or all possible combinations of one or more associated listed items.

[0045] It should be understood that although the terms "first", "second", "third", etc. may be used in this application to describe various information, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of this application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0046] With the rapid development of science and technology, autonomous driving technology has gradually penetrated into daily life. The task of autonomous driving trajectory planning is mainly to use map data, perception data, positioning data, vehicle status and other information to calculate a collision-free executable trajectory to ensure that the autonomous driving vehicle drives safely from the starting point to the destination and is as efficient as possible. The existing autonomous driving trajectory planning framework mainly adopts decoupled planning in the Frenet coordinate system and trajectory planning directly in three-dimensional space and time. The former uses s as the direction of the road centerline and l as the direction perpendicular to the road centerline. The latter plans the longitudinal s, lateral l, and time t at the same time, and takes the path and speed as variables of the problem at the same time to obtain an optimal combination. The latter directly considers maneuvers in space and time and directly handles planning problems in three-dimensional space, so the result is generally globally optimal. However, the speed of the three-dimensional trajectory planner is slower than the decoupled method, but it can better balance the relationship between calculation speed and solution quality.

[0047] In the process of obtaining the initial solution, a dynamic search is performed based on the sampling points. It is necessary to consider the kinematic constraints, curvature constraints, collision detection, traffic regulations constraints and other conditions between the sampling points to finally obtain the optimal solution. Since the search is performed in three-dimensional space, if uniform grid sampling is used on all axes, there will be thousands or even tens of thousands of condition checks during the search process, which consumes a lot of time. The initial solution will be used for convex optimization later, which will lead to a reduction in the frequency of the overall system communication and a delay in the response of the autonomous driving system.

[0048] Therefore, in the prior art, how to quickly search for an effective initial solution in three-dimensional space planning is an urgent problem to be solved.

[0049] Figure 1 is a flow chart of a key point sampling method for an initial solution of spatiotemporal joint planning shown in an embodiment of the present application, Figure 1 include:

[0050] Step S101, establish a space-time corridor during vehicle driving, the space-time corridor includes an s-axis generated in the direction of the road centerline, an l-axis perpendicular to the s-axis on the map plane, and a t-axis used to represent time, the s-axis, l-axis and t-axis are perpendicular to each other.

[0051] The sampling in step S101 relies on the space-time corridor generated based on the perception and prediction data of the obstacle. The perception data acquired by the space-time corridor includes the location, length and width information of the obstacle; and the predicted trajectory of the obstacle is given based on the perception data. Figure 2 This is a three-dimensional schematic diagram of a static obstacle in a space-time corridor according to an embodiment of the present application. Figure 3 This is a three-dimensional schematic diagram of a dynamic obstacle in a space-time corridor according to an embodiment of the present application. Figure 2As shown in , for static obstacles, the position is fixed at each moment, so relative to the fixed current lane line, there is continuous data on the axis, and the data at each time point is the same, so in Figure 2 The sl graph corresponding to each moment in is the same (the sl graph is a two-dimensional plane graph of the s axis and the l axis at the same moment). Figure 3 As shown in the figure, for dynamic obstacles, a graph is generated based on its predicted trajectory. Each point on the predicted trajectory has time and position information. Based on this information and the length and width of the obstacle, the obstacle relative to the current lane line at each time point can be obtained, which is often different at different times. Therefore, Figure 3 The obstacle coordinates in the sl map corresponding to each moment are different.

[0052] It should be noted that, in this embodiment, the s-axis, l-axis and t-axis of the space-time corridor are determined according to the position and posture of the vehicle, the direction of the s-axis is the same as the driving direction of the vehicle and the s-axis is parallel to the lane line, and the l-axis is perpendicular to the lane line. By establishing the space-time corridor, the vehicle will not find obstacles overlapping each other during driving.

[0053] Step S102, identifying obstacles and obtaining first position information of the obstacles in the space-time corridor.

[0054] In one embodiment, identifying obstacles includes: uniformly sampling in the s-axis direction of the space-time corridor, the l-axis direction of the space-time corridor, and the t-axis direction of the space-time corridor to identify obstacles.

[0055] It should be noted that, in this embodiment, the SL image in the space-time corridor is derived from obstacles in a high-precision map or a standard map. Since the accuracy of the standard map is lower than that of the high-precision map, the data density of the standard map is less than that of the high-precision map, that is, the sampling points included in the road section in the high-precision map are relatively dense, and the sampling points included in the road section in the standard map are relatively sparse, that is, for the same road section in the real world, the number of sampling points used to constitute the road section in the high-precision map is more than the number of sampling points used to constitute the road section in the standard map. Accordingly, after obtaining the obstacle information in the high-precision map, the sampling point density of the obstacle can be increased, including but not limited to resampling the key points.

[0056] Figure 4 is a schematic diagram of the first position information of the obstacle in the embodiment of the present application, such as Figure 4 As shown in , the sampling is performed uniformly on the time axis and continues along the s axis. For uniform sampling of time, the obstacle map cross section is taken at each time sampling point, that is, the first position information map of the relevant obstacle at that time point. Figure 4As shown in the figure, at time t1, the first position information of obstacle obj1 is bound1 in the figure below, the first position information of obstacle obj2 is bound2 in the figure below, and the first position information of obstacle obj3 is bound3 in the figure below. Due to the position, length, and width information of the obstacle, the first position information includes the coordinate information of the boundary pixel points of all obstacle images of the obstacle in the sl figure.

[0057] Step S103: expanding the first position information according to a preset safety distance to obtain second position information after the first position information is expanded, and generating a collection point range based on the second position information.

[0058] In one embodiment, the first position information is expanded according to a preset safety distance to obtain second position information after the first position information is expanded, including: expanding the first position information according to the preset safety distance; extending the first position information along the l-axis and in the direction away from the l-axis by a first distance respectively to obtain a first series of coordinates and a second series of coordinates, the first distance being the half-width length of the vehicle plus the preset safety distance; extending the first position information in the s-axis direction by a second distance to obtain a third series of coordinates, the second distance being the rear wheelbase length of the vehicle plus the preset safety distance; extending the first position information in the direction away from the s-axis by a third distance to obtain a fourth series of coordinates, the third distance being the front wheelbase length of the vehicle plus the safety distance.

[0059] Specifically, if Figure 5 As shown, Figure 5 This is a schematic diagram of the second position information of an obstacle shown in an embodiment of the present application. To ensure that the kinematic constraints are satisfied, sampling is still performed uniformly in the s direction. Sampling is performed in the l direction at a safe distance around the obstacle. Suppose there is an obstacle obj1, and its sl_boundary is expanded in width by half_width+safe_dis to obtain the first series of coordinates and the second series of coordinates; in front of the obstacle, back_over_hang+safe_dis is expanded to obtain the third series of coordinates; in the rear of the obstacle, front_over_hang+s afe_dis is expanded to obtain the fourth series of coordinates, and the expanded first position information is set as the second position information. Among them, half_width represents half the width of the vehicle, back_over_hang is the rear wheelbase of the vehicle, front_over_hang is the front wheelbase, and safe_dis represents the safe distance.

[0060] In one embodiment, generating a collection point range based on the second position information includes: obtaining a set of sampling variable distances, selecting the same sampling variable distance from the set, expanding the third series of coordinates and the fourth series of coordinates on the S axis, and generating a series of coordinates for the collection point range on the s axis; identifying the adjacent lane line range of the first series of coordinates and the second series of coordinates in a direction away from the obstacle, eliminating the coordinate range between the first series of coordinates and the second series of coordinates within the adjacent lane line range, and generating a series of coordinates for the collection point range on the l axis.

[0061] Specifically, Figure 6 As shown, Figure 6 This is a schematic diagram of the range of collection points shown in an embodiment of the present application. Figure 6 This is a screenshot of SL at a certain moment, where the solid line represents the lane line, the dotted line represents the center line of the road, and the part with diagonal lines represents the range of the collection points. Figure 6 The solid color filled part in is used to represent the second position information. For the acquisition point range, let its minimum value in the s-axis direction be min_s (i.e., the third series of coordinates) and its maximum value be max_s (i.e., the fourth series of coordinates), and its minimum value in the l-axis direction be min_l (i.e., the first series of coordinates) and its maximum value be max_l (i.e., the second series of coordinates). delta is used to represent the sampling variable distance. The series of coordinates on the s-axis of the acquisition point range are expressed using formulas (1) and (2), as shown below:

[0062] s1 = min_s-delta s Formula (1);

[0063] s2=max_s+delta s Formula (2);

[0064] In this embodiment, the third series of coordinates and the fourth series of coordinates are simultaneously expanded by the same sampling variable distance, so that sampling is performed around obstacles within the range of collection points.

[0065] In one embodiment, before generating the collection point range based on the second position information, it includes: identifying the second position information after expansion of multiple obstacles, and when the second position information after expansion of multiple obstacles overlaps, merging and calculating the second position information after expansion of the multiple obstacles; and generating the collection point range based on the merged second position information.

[0066] The embodiment of the present application further reduces the calculation process and error factors in the initial solution process by performing overlapping judgment on the second position information of multiple obstacles.

[0067] Step S104, within the range of the collection points, sampling is performed around the obstacle to obtain the target key points; when the range of the collection points is exceeded, sampling is performed on the center line of the road to obtain the target key points.

[0068] In one embodiment, within the range of collection points, sampling is performed around obstacles to obtain target key points, including: identifying the road centerline of the lane line, sampling the road centerline and the first series of coordinates and the second series of coordinates within the range of collection points; if the range of collection points exceeds the road boundary, then the sampling is abandoned within the range of collection points exceeding the road boundary.

[0069] In one embodiment, when the collection point range is exceeded, sampling is performed on the road centerline to obtain target key points, including: identifying the road centerline that exceeds the collection point range, and sampling equidistantly on the road centerline to obtain target key points.

[0070] Figure 7 This is a schematic diagram of a collection point of an obstacle shown in an embodiment of the present application. Figure 7 Among them, the solid line represents the lane line, the dotted line represents the road center line, the solid color filled part is used to represent the second position information, and the solid point is used to represent the target key point. Figure 7 In the example, the specific meaning of sampling around obstacles is that at sampling points outside the range of collection points, sampling is only performed on multiple road centerlines to preserve the topological relationship of lane lines. At sampling points within the range of collection points, sampling is only performed on multiple road centerlines within the range of collection points, and sampling is performed at the second position information of obstacles. If the road boundary is exceeded, sampling is abandoned. This is to provide a path to avoid obstacles, which can reduce the number of meaningless condition judgments caused by collision detection. The schematic diagram of the sampling points is as follows, where the blue area represents the area that needs to bypass obstacles.

[0071] Figure 8 This is a schematic diagram of collection points of multiple obstacles shown in an embodiment of the present application. Figure 8 The solid line represents the lane line, the dotted line represents the road center line, the solid color filled part is used to represent the second position information, and the solid point is used to represent the target key point. Figure 8 As shown, when there are multiple obstacles, the second position information of the multiple obstacles can be "ANDed" within the sampled range for each obstacle. That is, the second position information of each obstacle is combined and calculated, and then the collection point range is generated for the combined second position information.

[0072] In this embodiment, generating a collection point range for the second position information after the combined calculation includes: obtaining the first-class boundary coordinates and the second-class boundary coordinates of the combined second position information on the l-axis, obtaining the third-class boundary coordinates and the fourth-class boundary coordinates of the combined second position information on the s-axis, wherein the first-class boundary coordinates are smaller than the second-class boundary coordinates, and the third-class boundary coordinates are smaller than the fourth-class boundary coordinates. Obtaining a set of sampling variable distances, selecting the same sampling variable distance from the set, expanding the third-class coordinates and the fourth-class coordinates, and generating the collection point range on the s-axis class coordinates; identifying the adjacent lane line range of the first-class coordinates and the second-class coordinates in the direction away from the obstacle, eliminating the coordinate range between the first-class coordinates and the second-class coordinates in the adjacent lane line range, and generating the collection point range on the l-axis class coordinates.

[0073] In one embodiment, when the collection point range is exceeded, sampling is performed on the road center line to obtain the target key point, and the adjacent road section can be regarded as a line segment formed by orderly arrangement of more than 2 target key points. Accordingly, the target key point is a constituent element of the corresponding adjacent road section recorded in the standard map. Each target sampling point included in the adjacent road section can be the starting point, the end point and the middle point between the starting point and the end point of the adjacent road section. In addition, the arrangement order of each target key point included in the adjacent road section can also be formed according to the geographical location of each target key point in the real world. In the present embodiment, the road topology within the non-obstruction adoption range is retained by the target key point.

[0074] In one embodiment, after obtaining the target sampling point, the longitude and latitude information of the target key point can also be obtained, so as to facilitate the subsequent cancellation of resampling of the target key point according to the map solution. The resampling can be implemented by any applicable resampling method, such as the nearest neighbor interpolation (NNI) and bilinear interpolation (BI).

[0075] After completing the preliminary solution process of the target key points, this application establishes a three-dimensional space-time motion corridor generation model for the autonomous driving vehicle; based on the constraint condition that the autonomous driving vehicle has no collision with obstacles in the three-dimensional space-time motion corridor, the solution process of the autonomous driving path planning is realized.

[0076] Fig. 9 This is a diagram showing the effect of key point sampling for the initial solution of spatiotemporal joint planning in the embodiment of the present application. Fig. 9As shown in the figure, after sampling sl points on the cross section of all t-axis sampling points, all sparse key sampling points on slt are obtained. Compared with the grid map of the traditional space-time corridor, the number of sampling points is greatly reduced. Compared with the uniform grid sampling with the same sampling points in time t and longitudinal s, more than 4 times of time can be saved in the search for the initial solution during the space-time planning process. As the number of t sampling and s sampling points increases, more time is saved.

[0077] like Fig.10 As shown, the present invention also provides a key point sampling device for an initial solution of spatiotemporal joint planning in a second aspect of the present application, comprising:

[0078] A construction module 1001 is used to establish a space-time corridor during vehicle driving, wherein the space-time corridor includes an s-axis generated in the direction of the road centerline, an l-axis perpendicular to the s-axis on a map plane, and a t-axis used to represent time, wherein the s-axis, the l-axis and the t-axis are perpendicular to each other;

[0079] The identification module 1002 is used to identify obstacles and obtain first position information of the obstacles in the space-time corridor;

[0080] An expansion module 1003 is used to expand the first position information according to a preset safety distance to obtain second position information after the first position information is expanded, and generate a collection point range based on the second position information;

[0081] The sampling module 1004 is used to sample around obstacles within the range of collection points to obtain target key points; when the range of collection points is exceeded, sampling is performed on the center line of the road to obtain target key points.

[0082] In one embodiment, identifying obstacles includes: uniformly sampling in the s-axis direction of the space-time corridor, the l-axis direction of the space-time corridor, and the t-axis direction of the space-time corridor to identify obstacles.

[0083] In one embodiment, the first position information is expanded according to a preset safety distance to obtain second position information after the first position information is expanded, including: expanding the first position information according to the preset safety distance; extending the first position information along the l-axis and in the direction away from the l-axis by a first distance respectively to obtain a first series of coordinates and a second series of coordinates, the first distance being the half-width length of the vehicle plus the preset safety distance; extending the first position information in the s-axis direction by a second distance to obtain a third series of coordinates, the second distance being the rear wheelbase length of the vehicle plus the preset safety distance; extending the first position information in the direction away from the s-axis by a third distance to obtain a fourth series of coordinates, the third distance being the front wheelbase length of the vehicle plus the safety distance.

[0084] In one embodiment, generating a collection point range based on the second position information includes: obtaining a set of sampling variable distances, selecting the same sampling variable distance from the set, expanding the third series of coordinates and the fourth series of coordinates on the S axis, and generating a series of coordinates of the collection point range on the s axis; identifying the adjacent lane line range of the first series of coordinates and the second series of coordinates in the direction away from the obstacle, eliminating the coordinate range between the first series of coordinates and the second series of coordinates within the adjacent lane line range, and generating a series of coordinates of the collection point range on the l axis.

[0085] In one embodiment, before generating the collection point range based on the second position information, it includes: identifying the second position information after expansion of multiple obstacles, and when the second position information after expansion of multiple obstacles overlaps, merging and calculating the second position information after expansion of the multiple obstacles; and generating the collection point range based on the merged second position information.

[0086] In one embodiment, within the range of collection points, sampling is performed around obstacles to obtain target key points, including: identifying the road centerline of the lane line, sampling the road centerline and the first series of coordinates and the second series of coordinates within the range of collection points; if the range of collection points exceeds the road boundary, then the sampling is abandoned within the range of collection points exceeding the road boundary.

[0087] In one embodiment, when the collection point range is exceeded, sampling is performed on the road centerline to obtain target key points, including: identifying the road centerline that exceeds the collection point range, and sampling equidistantly on the road centerline to obtain target key points.

[0088] The technical solution provided by this application may have the following beneficial effects:

[0089] On the one hand, the present application scans the map obstacle map in a first direction to find the nearest obstacle in the first direction of the pixel point during the scanning process in the first direction; scans the map obstacle map according to the second direction to find the nearest obstacle in the second direction of the pixel point during the scanning process in the second direction; wherein the first direction and the second direction are opposite. The present application determines the nearest obstacle of the same pixel point by comparing the nearest obstacle in the first direction of the pixel point with the nearest obstacle in the second direction of the pixel point, and generates a candidate obstacle layer of the map obstacle map according to the two-dimensional coordinates of the nearest obstacle. Through scanning in the above two directions, the present application provides accurate obstacle location and distance information for path planning in autonomous driving.

[0090] On the other hand, the present application determines the target obstacle point of the vehicle during driving according to the vehicle body coordinate system, the preset pixel threshold and the candidate obstacle layer, and generates the target obstacle layer of the vehicle according to the target obstacle point. The present application restricts the conversion of the candidate obstacle layer by the preset pixel threshold, saving the memory resource overhead in the planning path.

[0091] Fig.11 It is a schematic diagram of the structure of an electronic device shown in an embodiment of the present application.

[0092] See also Fig.11 , the electronic device 1100 includes a memory 1110 and a processor 1120.

[0093] The processor 1120 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

[0094] The memory 1110 may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage. Among them, ROM may store static data or instructions required by the processor 1120 or other modules of the computer. The permanent storage may be a readable and writable storage device. The permanent storage may be a non-volatile storage device that does not lose the stored instructions and data even after the computer is powered off. In some embodiments, the permanent storage device uses a large-capacity storage device (such as a magnetic or optical disk, flash memory) as a permanent storage device. In other embodiments, the permanent storage device may be a removable storage device (such as a floppy disk, optical drive). The system memory may be a readable and writable storage device or a volatile readable and writable storage device, such as a dynamic random access memory. The system memory may store some or all instructions and data required by the processor at run time. In addition, the memory 1110 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (such as DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and disks and / or optical disks may also be used. In some embodiments, the memory 1110 may include a readable and / or writable removable storage device, such as a laser disc (CD), a read-only digital versatile disc (such as a DVD-ROM, a double-layer DVD-ROM), a read-only Blu-ray disc, an ultra-density optical disc, a flash memory card (such as an SD card, a mini SD card, a Micro-SD card, etc.), a magnetic floppy disk, etc. The computer-readable storage medium does not include carrier waves and transient electronic signals transmitted wirelessly or wired.

[0095] The memory 1110 stores executable codes, and when the executable codes are processed by the processor 1120 , the processor 1120 can execute part or all of the methods described above.

[0096] The above-mentioned electronic device can exist in the form of an autonomous driving domain controller, wherein the autonomous driving domain controller is communicated with various vehicle-mounted sensors (including but not limited to lidar, camera, millimeter-wave radar, ultrasonic radar, inertial measurement unit, wheel speed meter, etc.) loaded on the mobile tool, obtains environmental perception data and vehicle speed information through these sensors, extracts obstacle information and generates the position information of the mobile tool based on the acquired environmental perception data and vehicle speed information, and performs path planning based on the position information and obstacle information.

[0097] In addition, the method according to the present application may also be implemented as a computer program or a computer program product, which includes computer program code instructions for executing some or all of the steps in the above method of the present application.

[0098] Alternatively, the present application can also be implemented as a computer-readable storage medium (or non-transitory machine-readable storage medium or machine-readable storage medium) on which executable code (or computer program or computer instruction code) is stored. When the executable code (or computer program or computer instruction code) is executed by a processor of an electronic device (or server, etc.), the processor executes part or all of the steps of the above-mentioned method according to the present application.

[0099] The present application also provides a mobile tool, which includes Fig.11 Example electronic devices.

[0100] The "mobile tool" referred to in this application can be a vehicle or robot with the following functions: (1) passenger function, such as a family car, a bus, etc.; (2) cargo function, such as an ordinary truck, a van, a trailer, a closed truck, a tank truck, a flatbed truck, a container truck, a dump truck, a truck with a special structure, etc.; (3) tool function, such as a logistics distribution vehicle, an automatic guided vehicle AGV, a patrol car, a crane, a crane, an excavator, a bulldozer, a forklift, a road roller, a loader, an off-road engineering vehicle, an armored engineering vehicle, a sewage treatment vehicle, a sanitation vehicle, a vacuum truck, a floor scrubber, a watering truck, a sweeping robot, a food delivery robot, a shopping guide robot, a lawn mower, a golf cart, etc.; (4) entertainment function, such as an entertainment vehicle, an amusement park automatic driving device, a balance car, etc.; (5) special rescue function, such as a fire truck, an ambulance, a power repair vehicle, an engineering rescue vehicle, etc.

[0101] Vehicles or robots with the above functions include but are not limited to vehicles or robot equipment with L0-L5 autonomous driving technology levels as established by the Society of Automotive Engineers International (SAE International) or the Chinese national standard "Automotive Driving Automation Classification".

[0102] The embodiments of the present application have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A key point sampling method for initial solution of spatiotemporal joint planning, characterized in that: include: Establishing a space-time corridor during vehicle driving, wherein the space-time corridor includes an s-axis generated in the direction of the road centerline, an l-axis perpendicular to the s-axis on a map plane, and a t-axis for representing time, wherein the s-axis, the l-axis, and the t-axis are perpendicular to each other; Identify obstacles and obtain first position information of the obstacles in the space-time corridor; Expanding the first location information according to a preset safety distance to obtain second location information after the first location information is expanded, and generating a collection point range based on the second location information; Within the range of the collection points, sampling is performed around the obstacle to obtain target key points; When the range of the acquisition points is exceeded, sampling is performed on the road centerline to obtain the target key points.

2. The method according to claim 1, characterized in that The obstacle identification includes: Obstacles are identified by uniformly sampling in the s-axis direction of the space-time corridor, the l-axis direction of the space-time corridor, and the t-axis direction of the space-time corridor.

3. The method according to claim 1, characterized in that The step of expanding the first location information according to the preset safety distance to obtain second location information after the first location information is expanded includes: Expanding the first location information according to a preset safety distance; Extend the first position information by a first distance along the direction of the l-axis and in the direction away from the l-axis respectively to obtain a first series of coordinates and a second series of coordinates, wherein the first distance is the half-width length of the vehicle plus a preset safety distance; The first position information is extended by a second distance in the s-axis direction to obtain a third series of coordinates , The second distance is the length of the rear wheelbase of the vehicle plus a preset safety distance; The first position information is extended a third distance away from the s-axis direction to obtain a fourth series of coordinates, and the third distance is the length of the front wheelbase of the vehicle plus a safety distance.

4. The method according to claim 3, characterized in that: The generating a collection point range based on the second position information includes: Obtaining a set of sampling variable distances, selecting the same sampling variable distance from the set, expanding the third series of coordinates and the fourth series of coordinates on the S axis, and generating a series of coordinates with a collection point range on the S axis; Identify the adjacent lane line range of the first series of coordinates and the second series of coordinates in the direction away from the obstacle, eliminate the coordinate range between the first series of coordinates and the second series of coordinates in the adjacent lane line range, and generate a series of coordinates of the collection point range on the l-axis.

5. The method according to claim 4, characterized in that Before generating the collection point range based on the second location information, the method includes: Identify the expanded second position information of multiple obstacles, and when the expanded second position information of multiple obstacles overlap, combine and calculate the expanded second position information of the multiple obstacles; The collection point range is generated based on the combined second position information.

6. The method according to claim 4, characterized in that The method of sampling around the obstacle within the range of the collection points to obtain the target key points includes: Identify the road centerline of the lane line, and sample the road centerline and the first series of coordinates and the second series of coordinates within the collection point range; If the range of the collection points exceeds the road boundary, sampling is abandoned at the range of the collection points exceeding the road boundary.

7. The method according to claim 1, characterized in that When the range of the collection points is exceeded, sampling is performed on the center line of the road to obtain the target key points, including: The road centerline beyond the range of the collection points is identified, and sampling is performed at equal intervals on the road centerline to obtain the target key points.

8. A key point sampling device for initial solution of spatiotemporal joint planning, characterized in that: include: A construction module is used to establish a space-time corridor during vehicle driving, wherein the space-time corridor includes an s-axis generated in the direction of the road centerline, an l-axis perpendicular to the s-axis on a map plane, and a t-axis used to represent time, wherein the s-axis, the l-axis and the t-axis are perpendicular to each other; An identification module, used to identify an obstacle and obtain first position information of the obstacle in the space-time corridor; An expansion module, configured to expand the first location information according to a preset safety distance to obtain second location information after the first location information is expanded, and generate a collection point range based on the second location information; A sampling module, used for sampling around the obstacle within the range of the collection points to obtain target key points; When the range of the acquisition points is exceeded, sampling is performed on the road centerline to obtain the target key points.

9. An electronic device, characterized in that: include: processor; as well as A memory having executable codes stored thereon, which, when executed by the processor, causes the processor to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: An executable code is stored thereon, and when the executable code is executed by a processor of an electronic device, the processor is caused to execute the method as claimed in any one of claims 1 to 7.

11. A mobile tool, characterized in that: Comprising the electronic device as claimed in claim 10.

Citation Information

Cited By

  • Path planning method and device

    CN120970657A