Path planning method and device, unmanned vehicle and storage medium

Abstract

The application discloses a path planning method and device, an unmanned vehicle and a storage medium, wherein the method comprises the following steps: acquiring an initial reference path of a target vehicle and initial and terminal collision points of an obstacle object; determining a local planning interval of the target vehicle based on the initial and terminal collision points, wherein the local planning interval extends along an S-axis direction of a road coordinate system, and the local planning interval intersects with the initial reference path at a planning starting point and a planning terminal point; determining a first obstacle avoidance position of the target vehicle at an initial collision moment based on the initial collision point, and determining a second obstacle avoidance position of the target vehicle at a terminal collision moment based on the terminal collision point; determining an obstacle avoidance reference path of the target vehicle in the local planning interval based on the planning starting point, the planning terminal point and the first and second obstacle avoidance positions of the target vehicle; and planning a reference path of the target vehicle based on the obstacle avoidance reference path. In this way, the reference path can be quickly planned.

Description

Technical Field

[0001] This application belongs to the field of autonomous driving technology, specifically relating to a path planning method and device, an unmanned vehicle, and a storage medium. Background Technology

[0002] In recent years, autonomous driving technology has developed rapidly. Its goal is usually to control the vehicle to travel autonomously along the road, to reach the destination as quickly as possible while ensuring the safety of the vehicle itself, and to ensure that it does not pose a direct or indirect threat to the safety of other road users.

[0003] To achieve the above goals, autonomous driving software requires several key systems, one of which is the path planning system. The goal of path planning is to plan a reasonable reference path within the vehicle's driving area. This reference path should meet various set constraints and ensure that the vehicle will not collide with any obstacles within the driving area while traveling along the reference path.

[0004] The information disclosed in this background section is intended only to enhance the understanding of the overall background of this application and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this application is to provide a path planning method for solving the problem of how to implement path planning for autonomous driving.

[0006] To achieve the above objectives, this application provides a path planning method, the method comprising:

[0007] Obtain the initial reference path of the target vehicle and the initial and final collision points of the obstacle objects;

[0008] Based on the initial collision point and the final collision point, a local planning interval for the target vehicle is determined, wherein the local planning interval extends along the S-axis direction of the road coordinate system, and the local planning interval intersects with the initial reference path at the planning start point and the planning end point.

[0009] The first obstacle avoidance position of the target vehicle at the time of the initial collision is determined based on the initial collision point, and the second obstacle avoidance position of the target vehicle at the time of the final collision is determined based on the final collision point.

[0010] Based on the planning start point, planning end point, and the first and second obstacle avoidance positions of the target vehicle, determine the obstacle avoidance reference path of the target vehicle in the local planning interval;

[0011] Based on the obstacle avoidance reference path, a reference path for the target vehicle is planned.

[0012] In one embodiment, determining the local planning interval of the target vehicle based on the initial collision point and the final collision point specifically includes:

[0013] Based on preset rules, the obstacle avoidance adjustment time is divided into a first adjustment time and a second adjustment time;

[0014] A first adjustment value is determined based on the first adjustment duration and the speed of the target vehicle, and a second adjustment value is determined based on the second adjustment duration and the speed of the target vehicle.

[0015] On the initial reference path, the planning start point of the local planning interval is determined by adjusting the first adjustment value in reverse from the initial collision point, and the planning end point of the local planning interval is determined by adjusting the second adjustment value in forward from the final collision point.

[0016] In one embodiment, the obstacle object includes a first obstacle object and a second obstacle object;

[0017] Based on the initial collision point and the final collision point, the local planning interval of the target vehicle is determined, specifically including:

[0018] Based on the initial collision point and final collision point of the initial reference path and the first obstacle object, a first sub-local planning interval of the target vehicle is determined, and based on the initial collision point and final collision point of the initial reference path and the second obstacle object, a second sub-local planning interval of the target vehicle is determined, wherein the first sub-local planning interval and the second sub-local planning interval extend along the S-axis direction of the road coordinate system.

[0019] Determine whether the first sub-local planning interval and the second sub-local planning interval include overlapping intervals; if so,

[0020] The first sub-local planning interval and the second sub-local planning interval are merged into the local planning interval.

[0021] In one embodiment, determining the first obstacle avoidance position of the target vehicle at the initial collision moment based on the initial collision point specifically includes:

[0022] Determine the first target edge point corresponding to the obstacle object and the initial collision point, wherein, in the road coordinate system, the L-axis coordinate of the first target edge point takes an extreme value and is equal to the S-axis coordinate of the initial collision point;

[0023] Based on the first target edge point and the preset vehicle body expansion coefficient, the first obstacle avoidance position of the target vehicle at the initial collision moment is determined;

[0024] And / or,

[0025] Determining the second obstacle avoidance position of the target vehicle at the moment of the final collision based on the terminal collision point specifically includes:

[0026] Determine the second target edge point corresponding to the obstacle object and the final collision point, wherein, in the road coordinate system, the L-axis coordinate of the second target edge point takes an extreme value and is equal to the S-axis coordinate of the final collision point;

[0027] Based on the second target edge point and the preset vehicle body expansion coefficient, the second obstacle avoidance position of the target vehicle at the moment of final collision is determined.

[0028] In one embodiment, determining the first obstacle avoidance position of the target vehicle at the initial collision moment based on the first target edge point and a preset vehicle body expansion coefficient specifically includes:

[0029] Based on the first target edge point and the expanded vehicle width, the L-axis coordinate range of the first obstacle avoidance position in the road coordinate system is determined, wherein the expanded vehicle width is determined by the expansion of the target vehicle width by the preset vehicle body expansion coefficient.

[0030] The S-axis coordinate range of the first obstacle avoidance position in the road coordinate system is determined based on the first target edge point and the reference vehicle length of the target vehicle, wherein the S-axis coordinate of the first target edge point is located within the S-axis coordinate range of the first obstacle avoidance position.

[0031] And / or,

[0032] Based on the second target edge point and a preset vehicle body expansion coefficient, the second obstacle avoidance position of the target vehicle at the moment of final collision is determined, specifically including:

[0033] Based on the second target edge point and the expanded vehicle width, the L-axis coordinate range of the second obstacle avoidance position in the road coordinate system is determined, wherein the expanded vehicle width is determined by the expansion of the target vehicle width by the preset vehicle body expansion coefficient;

[0034] The S-axis coordinate range of the second obstacle avoidance position in the road coordinate system is determined based on the second target edge point and the reference vehicle length of the target vehicle, wherein the S-axis coordinate of the second target edge point is located within the S-axis coordinate range of the second obstacle avoidance position.

[0035] In one embodiment, determining the S-axis coordinate interval of the first obstacle avoidance position in the road coordinate system based on the first target edge point and the reference vehicle length of the target vehicle specifically includes:

[0036] In the road coordinate system, the center point of the first obstacle avoidance position is determined, wherein the S-axis coordinate of the center point of the first obstacle avoidance position is equal to the S-axis coordinate of the edge point of the first target.

[0037] Based on the center point of the first obstacle avoidance position and the reference vehicle length, determine the S-axis coordinate range of the first obstacle avoidance position;

[0038] And / or,

[0039] The S-axis coordinate range of the second obstacle avoidance position in the road coordinate system is determined based on the second target edge point and the reference vehicle length of the target vehicle, specifically including:

[0040] In the road coordinate system, the center point of the second obstacle avoidance position is determined, wherein the S-axis coordinate of the center point of the second obstacle avoidance position is equal to the S-axis coordinate of the edge point of the second target.

[0041] Based on the center point of the second obstacle avoidance position and the reference vehicle length, the S-axis coordinate range of the second obstacle avoidance position is determined.

[0042] In one embodiment, the reference vehicle length is twice the width of the expanded vehicle.

[0043] In one embodiment, the first obstacle avoidance position is used to define a first heading of the target vehicle at the initial collision time, and the second obstacle avoidance position is used to define a second heading of the target vehicle at the final collision time;

[0044] Based on the planning start point, planning end point, and the first and second obstacle avoidance positions of the target vehicle, an obstacle avoidance reference path for the target vehicle within the local planning interval is determined, specifically including:

[0045] In the road coordinate system, a first reference line segment is determined that passes through the center point of the first obstacle avoidance position and extends along the first heading, and a second reference line segment is determined that passes through the center point of the second obstacle avoidance position and extends along the second heading, wherein the lengths of the first reference line segment and the second reference line segment are equal to the reference vehicle length.

[0046] By sequentially connecting the planning starting point, the first reference line segment, the second reference line segment, and the planning ending point, the obstacle avoidance reference path of the target vehicle in the local planning interval is obtained.

[0047] In one embodiment, planning a reference path for the target vehicle based on the obstacle avoidance reference path specifically includes:

[0048] Based on the obstacle avoidance reference path, a fusion reference path is constructed, wherein the fusion reference path is formed by connecting the initial reference path after removing the portion within the local planning interval and the obstacle avoidance reference path;

[0049] Based on the fused reference path, a reference path for the target vehicle is planned.

[0050] In one embodiment, planning a reference path for the target vehicle based on the fused reference path specifically includes:

[0051] In the road coordinate system, a sub-loss function is constructed based on the L-axis coordinate interval of the obstacle avoidance reference path in the local planning interval;

[0052] Based on the sub-loss function and the loss function of the preset path planning model, a comprehensive loss function is constructed;

[0053] With the goal of minimizing the comprehensive loss function, the reference path of the target vehicle is planned without constraints based on the preset path planning model and the fused reference path.

[0054] In one embodiment, the method further includes:

[0055] Confirm whether the maximum curvature on the reference path exceeds a preset curvature threshold; if so,

[0056] The length of the local planning interval is expanded to replan the reference path of the target vehicle.

[0057] In one embodiment, the method further includes:

[0058] Detect whether there is a collision point between the reference path and the obstacle object; if so,

[0059] Expand the L-axis coordinate range, or the L-axis and S-axis coordinate range, of the first and second obstacle avoidance positions in the road coordinate system to replan the reference path of the target vehicle.

[0060] This application also provides a path planning device, comprising:

[0061] The acquisition module is used to acquire the initial reference path of the target vehicle and the initial and final collision points of the obstacle objects;

[0062] The first determining module is used to determine the local planning interval of the target vehicle based on the initial collision point and the final collision point, wherein the local planning interval extends along the S-axis direction of the road coordinate system, and the local planning interval intersects the initial reference path at the planning start point and the planning end point.

[0063] The second determining module is used to determine the first obstacle avoidance position of the target vehicle at the time of the initial collision based on the initial collision point, and to determine the second obstacle avoidance position of the target vehicle at the time of the final collision based on the final collision point.

[0064] The third determining module is used to determine the obstacle avoidance reference path of the target vehicle in the local planning interval based on the planning start point, planning end point, and the first obstacle avoidance position and the second obstacle avoidance position of the target vehicle;

[0065] The planning module is used to plan a reference path for the target vehicle based on the obstacle avoidance reference path.

[0066] This application also provides an unmanned vehicle, including:

[0067] At least one processor; and

[0068] The memory stores instructions that, when executed by the at least one processor, cause the at least one processor to perform the path planning method as described above.

[0069] This application also provides a machine-readable storage medium storing executable instructions that, when executed, cause the machine to perform the path planning method described above.

[0070] Compared with the prior art, the path planning method of this application determines a local planning interval and determines an obstacle avoidance reference path based on the first obstacle avoidance position and the second obstacle avoidance position of the target vehicle within the local planning interval. In this way, when planning the reference path of the target vehicle, the obstacle avoidance reference path can be used as a reference, thereby achieving local optimization when replanning the reference path.

[0071] On the other hand, the reference path for planning is iterated in two layers through curvature detection and collision detection. In this way, the solution of the reference path can be unconstrained planning, which requires less computing power and can meet the real-time requirements of the algorithm.

[0072] On the other hand, by constructing a fusion reference path by removing a portion of the local planning interval based on the obstacle avoidance reference path and the initial reference path, and by planning a reference path based on the fusion reference path, the impact on the original initial reference path outside the local planning interval can be minimized. Attached Figure Description

[0073] Figure 1 This is an application scenario diagram of a path planning method according to an embodiment of this application;

[0074] Figure 2 This is a flowchart of a path planning method according to an embodiment of this application;

[0075] Figure 3 This is a scene diagram of a local planning interval for a single obstacle object in a path planning method according to an embodiment of this application;

[0076] Figure 4 This is a scene diagram of a local planning interval when there are two obstacle objects in a path planning method according to an embodiment of this application;

[0077] Figure 5This is a scene diagram illustrating the determination of the first obstacle avoidance position and the second obstacle avoidance position in a path planning method according to an embodiment of this application.

[0078] Figure 6 This is a scene diagram of determining the first obstacle avoidance position in a path planning method according to an embodiment of this application;

[0079] Figure 7 and Figure 8 This is a scene diagram illustrating the determination of different first and second obstacle avoidance positions in a path planning method according to an embodiment of this application.

[0080] Figure 9 This is a scene diagram of determining an obstacle avoidance reference path in a path planning method according to an embodiment of this application;

[0081] Figure 10 This is a scene diagram of determining a reference path in a path planning method according to an embodiment of this application;

[0082] Figure 11 A block diagram of a path planning device according to an embodiment of this application;

[0083] Figure 12 This is a hardware structure diagram of an unmanned vehicle according to an embodiment of this application. Detailed Implementation

[0084] The present application will now be described in detail with reference to the embodiments shown in the accompanying drawings. However, these embodiments do not limit the present application, and any structural, methodological, or functional modifications made by those skilled in the art based on these embodiments are included within the protection scope of the present application.

[0085] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “corresponding to,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0086] Before introducing the embodiments of this application, the basic technologies and some technical terms involved in the embodiments of this application will be explained illustratively:

[0087] Autonomous driving refers to the ability to guide and make decisions regarding vehicle operation without requiring a driver to perform physical driving maneuvers, thus enabling the vehicle to drive safely. Autonomous driving technology typically includes high-precision mapping, environmental perception, behavioral decision-making, path planning, and motion control.

[0088] Autonomous driving systems: Systems that enable different levels of autonomous driving functions in vehicles, such as driver assistance systems (L2), high-speed autonomous driving systems requiring human supervision (L3), and highly / fully autonomous driving systems (L4 / L5).

[0089] Intelligent Traffic Systems (ITS), also known as Intelligent Transportation Systems, effectively integrate advanced science and technology (information technology, computer technology, data communication technology, sensor technology, electronic control technology, automatic control theory, operations research, and artificial intelligence) into transportation, service control, and vehicle manufacturing. This strengthens the connection between vehicles, roads, and users, thereby forming a comprehensive transportation system that ensures safety, improves efficiency, enhances the environment, and conserves energy.

[0090] Intelligent Vehicle Infrastructure Cooperative Systems (IVICS) represent a future direction in intelligent transportation systems. IVICS utilizes advanced wireless communication and next-generation internet technologies to implement comprehensive, real-time dynamic information exchange between vehicles and between vehicles and infrastructure. Based on the collection and fusion of dynamic traffic information across all times and spaces, it enables proactive vehicle safety control and collaborative road management, fully realizing effective coordination between people, vehicles, and roads. This ensures traffic safety, improves traffic efficiency, and ultimately forms a safe, efficient, and environmentally friendly road transportation system.

[0091] Frenet coordinate system: also called road coordinate system, with the vehicle's starting position as the origin, the coordinate axes are perpendicular to each other, divided into the S-axis direction (that is, along the tangent direction of the road reference line, called the lateral direction) and the L-axis direction (that is, the current normal direction of the reference line, called the longitudinal direction), and the coordinates are represented as (S, L).

[0092] The path planning method according to embodiments of the present application can be applied to intelligent transportation systems, such as scenarios including driving navigation, logistics transportation, takeout delivery, and online ride-hailing. Based on the solutions of the embodiments of the present application, a reference path can be secondarily planned for an obstacle object detected ahead during vehicle traveling, which requires low computing power, and the planning speed can meet real-time requirements. According to different application scenarios, the path planning method provided by the embodiments of the present application can be applied to automatic driving systems of automatic driving vehicles, including levels L2, L3, L4 and above.

[0093] Referring Figure 1 , an application scenario of the path planning method provided by the embodiments of the present application is taken as an example. The vehicle can be manually driven by a user, or can automatically drive with the assistance of the intelligent driving system of the vehicle. In both manual driving and automatic driving processes, the terminal can collect scene information based on sensors, lidars, cameras, millimeter-wave radars, navigation systems, positioning systems, high-precision maps and the like, and provide decision basis information for vehicle control. The terminal can be the vehicle driven by the user, or an intelligent vehicle-mounted device / module on the vehicle, or a desktop computer, a notebook computer, a smart phone and a tablet computer configured on the vehicle when the user drives the vehicle, as well as a portable wearable device carried by the user, etc.

[0094] Referring Figure 2 , an embodiment of the path planning method of the present application is described. In this embodiment, the method includes:

[0095] S11, acquiring an initial reference path of the target vehicle and an initial collision point and a final collision point of the obstacle object.

[0096] In various embodiments of the present application, the initial reference path of the target vehicle can be planned by a variety of path planning methods. Different path planning methods may involve different preset path planning models, such as spline curve models, Bezier curve models, etc., and a constrained path solution is performed based on a set loss function.

[0097] In an exemplary embodiment, the initial reference path of the target vehicle is fitted by using multi-segment fourth-order spline curves. The i-th segment of the fourth-order spline curve can be expressed as:

[0098]

[0099] Wherein, the domain of the segmented fourth-order spline curve is from t i to t i+1 , which is the coordinate length on the S-axis of the time span h i in the road coordinate system, it can be known that h i =t i+1 -t i (that is, △t).

[0100] Set x i (t)=f j (t), x i and x i+1 Let L-axis coordinates be the starting and ending positions of the segments of the fourth-order spline curve, respectively. Differentiating equation (1) with respect to time order yields t. i The derivatives of the time-time curve along the s-axis are as follows:

[0101]

[0102]

[0103]

[0104]

[0105]

[0106] in, They represent x respectively i In t i The first, second, third, and fourth derivatives at that point, that is, Corresponding to the target vehicle at t i The velocity, acceleration, impact, and oscillation at the point.

[0107] Furthermore, formula (2) can be expressed in matrix form as follows:

[0108]

[0109] Similarly, t i+1 The derivatives of the time-time curve along the s-axis are as follows:

[0110]

[0111]

[0112]

[0113]

[0114]

[0115] Since the piecewise segments of the fourth-order spline curve are third-order differentiable at each connection point, that is:

[0116]

[0117] In mathematics, if the starting state variable (i.e., the third-order internal derivative) of a spline curve is known, and the fourth-order derivative with respect to the parameter variable at that starting point is obtained, then the analytical expression of the spline curve and the state variable at the endpoint of the spline curve can be uniquely determined. Therefore, the above formula (5) can be expressed by a matrix model as follows:

[0118]

[0119] Model observations are established based on formula (6). Control quantity The discrete state-space equation for the spline curve is:

[0120]

[0121] in:

[0122]

[0123] For path planning, the overall state space is: in Given the initial state of the target vehicle, the control space is: Because it is known and The following can be determined Therefore, there is no need for further action. Optimize.

[0124] As an example, the constraints for solving the model can be, for example, the boundary of the current driving path of the target vehicle, and the loss function can be, for example, the sum of the squares of the fourth derivatives of each spline curve segment. A smooth initial reference path is obtained by quadratic programming. This application does not limit this.

[0125] The initial reference path solutions given in the examples above typically include soft constraints, meaning the solved initial reference path carries the risk of colliding with obstacles; or, during the target vehicle's journey, sensors detect obstacles on the initial reference path ahead. In these scenarios, re-executing a constrained solution for the initial reference path would not meet the algorithm's real-time requirements.

[0126] Based on this, the embodiments of this application propose to optimize the path in a certain interval for the initial reference path that may cause collisions, and to reduce the computing power requirement while enhancing the real-time performance of the algorithm by solving the reference path without constraints.

[0127] Coordination Figure 3Specifically, after obtaining the initial reference path of the target vehicle, potential obstacles are identified based on this initial reference path, and the initial collision point Sin and the final collision point Sout between the initial reference path and the obstacle are obtained. Typically, at the initial collision point Sin, the target vehicle and the obstacle may have a frontal corner collision, while at the final collision point Sout, the target vehicle and the obstacle may have a rearal corner collision.

[0128] S12. Based on the initial collision point and the final collision point, determine the local planning interval of the target vehicle.

[0129] Coordination Figure 3 The local planning interval extends along the S-axis of the road coordinate system. Understandably, the L-axis coordinate range of the local planning interval in the road coordinate system can be determined by the lane width currently occupied by the target vehicle. The local planning interval can be used to define the main obstacle avoidance area for the target vehicle, and each local planning interval should include at least one obstacle. One of the objectives of this application is to at least locally optimize the reference path within the local planning interval. After entering the local planning interval, the target vehicle can avoid obstacles within the local planning interval by following the reference path and changing its heading angle.

[0130] It should be noted that in different embodiments of this application, the reference path planning performed by the target vehicle to avoid obstacles may be limited to a local planning area or may involve replanning the entire reference path. When replanning the entire reference path, the target vehicle may change its heading angle before entering the local planning area and after leaving the local planning area in order to avoid obstacles within the local planning area; this application does not impose any restrictions on this.

[0131] In the road coordinate system, the S-axis coordinates of the initial collision point Sin and the final collision point Sout should lie within the local planning interval. That is, the minimum value Smin of the local planning interval should be less than the S-axis coordinate of the initial collision point Sin, and the maximum value Smax of the local planning interval should be greater than the S-axis coordinate of the final collision point Sout.

[0132] In one embodiment, a preset adjustment value can be adjusted backward from the initial collision point Sin and forward from the final collision point Sout along the initial reference path to obtain a local planning interval. Here, "backward" and "forward" refer to the direction the target vehicle travels along the initial reference path; the direction same as this direction is considered "forward," and the opposite direction is considered "backward." Furthermore, the preset adjustment values ​​for the backward and forward adjustments can be the same or different.

[0133] Coordination Figure 3In one embodiment, an obstacle avoidance adjustment time can be preset, and this time can be divided into a first adjustment time and a second adjustment time based on preset rules. Then, a first adjustment value can be determined based on the first adjustment time and the target vehicle's speed, and a second adjustment value can be determined based on the second adjustment time and the target vehicle's speed. Finally, on the initial reference path, the first adjustment value is adjusted in reverse from the initial collision point Sin to determine the planning start point Sinit of the local planning interval, and the second adjustment value is adjusted in forward from the final collision point Sout to determine the planning end point Sfinal of the local planning interval.

[0134] The obstacle avoidance adjustment time setting can take into account factors such as vehicle speed and road curvature. Higher speeds and curvatures allow for a longer preset obstacle avoidance adjustment time, while lower speeds and curvatures allow for a shorter preset time. For example, the obstacle avoidance adjustment time (ttc) is preset to 1.2 seconds, divided into a first adjustment time of 0.6 seconds and a second adjustment time of 0.6 seconds. Then, the planning start point Sinit = Sin - 0.6 * v, the planning end point Sfinal = Sout + 0.6 * v, and the local planning interval Sop = Sfinal - Sinit. Alternatively, the 1.2-second obstacle avoidance adjustment time can be divided into a first adjustment time of 0.8 seconds and a second adjustment time of 0.4 seconds. Then, the planning start point Sinit = Sin - 0.8 * v and the planning end point Sfinal = Sout + 0.4 * v.

[0135] Coordination Figure 4 In some scenarios, there may be more than one obstacle object. Here, we will take two obstacle objects, the first obstacle object and the second obstacle object, as an example to explain the determination of the local planning interval in this case.

[0136] In this embodiment, the initial reference path has an initial collision point Sin1 and a final collision point Sout1 with the first obstacle object, and an initial collision point Sin2 and a final collision point Sout2 with the second obstacle object. Similarly, based on the initial collision point Sin1 and the final collision point Sout1, the first sub-local planning interval [Sinit 1, Sfinal1] of the target vehicle can be determined, and based on the initial collision point Sin2 and the final collision point Sout2, the second sub-local planning interval [Sinit 2, Sfinal 2] of the target vehicle can be determined. The method for determining the first and second sub-local planning intervals can be partially or entirely referred to the above embodiment, and will not be repeated here.

[0137] Further, determine whether the first and second sub-local planning intervals include overlapping intervals; if so, merge the first and second sub-local planning intervals into a single local planning interval. The criterion for determining whether the first and second sub-local planning intervals include overlapping intervals can be the positional relationship between the planning endpoint Sfinal 1 of the first sub-local planning interval and the planning starting point Sinit 2 of the second sub-local planning interval. If the planning endpoint Sfinal 1 is located in the positive direction of the planning starting point Sinit 2 on the initial reference path, then the first and second sub-local planning intervals include overlapping intervals; otherwise, they do not.

[0138] Similarly, when the initial reference path collides with three or more possible obstacle objects, whether to merge the sub-local planning intervals corresponding to different obstacle objects can be referred to the scenario of two obstacle objects mentioned above, which will not be elaborated here.

[0139] S13. Determine the first obstacle avoidance position of the target vehicle at the time of the initial collision based on the initial collision point, and determine the second obstacle avoidance position of the target vehicle at the time of the final collision based on the final collision point.

[0140] Coordination Figure 5 The first obstacle avoidance position and the second obstacle avoidance position can be "limit obstacle avoidance positions" used to determine the target vehicle at the initial collision time T1 and the final collision time T2, respectively. That is, this application assumes that at the initial collision time, if the target vehicle is in the first obstacle avoidance position, it can avoid colliding with the front corner of the obstacle; and at the final collision time, if the target vehicle is in the second obstacle avoidance position, it can avoid colliding with the rear corner of the obstacle.

[0141] The first obstacle avoidance position and the second obstacle avoidance position can be a reference point, and for example, correspond to the projection center of the target vehicle to determine the position of the target vehicle; or, the outlines of the first obstacle avoidance position and the second obstacle avoidance position can be "approximations" of the target vehicle's projection on the road coordinate system to determine the position of the target vehicle.

[0142] Based on the above assumptions, for example, the outlines of the first and second obstacle avoidance positions can be considered as rectangles, whose projection on the road coordinate system can cover the target vehicle. It should be noted that the shape and size of the outlines of the first and second obstacle avoidance positions can be adaptively adjusted according to the needs of the path planning method of this application. The term "approximation" mentioned above is not intended to limit the shape and size of the outlines of the first and second obstacle avoidance positions. For ease of description, the following embodiments use rectangles as an example to specifically illustrate the path planning method provided by this application.

[0143] In this embodiment, the first obstacle avoidance position of the target vehicle at the initial collision moment can be determined by identifying the first target edge point C1 corresponding to the obstacle object and the initial collision point, and based on the first target edge point C1 and a preset vehicle body expansion coefficient. Similarly, the second obstacle avoidance position of the target vehicle at the final collision moment can be determined by identifying the second target edge point C2 corresponding to the obstacle object and the final collision point, and based on the second target edge point C2 and a preset vehicle body expansion coefficient.

[0144] The preset vehicle body expansion coefficient can be used to expand the outlines of the first and second obstacle avoidance positions, thereby causing the outlines of the first and second expansion positions to expand relative to the actual projection of the target vehicle. In this way, when planning a reference path using the first and second obstacle avoidance positions, the target vehicle can be more likely to avoid obstacles.

[0145] In this embodiment, the L-axis coordinate of the first target edge point is an extreme value and equal to the S-axis coordinate of the initial collision point. This means that, in the direction of travel of the target vehicle, it is necessary to ensure that at least at the initial collision moment, the first obstacle avoidance position does not interfere with the first target edge point. Similarly, the L-axis coordinate of the second target edge point is an extreme value and equal to the S-axis coordinate of the final collision point. In the direction of travel of the target vehicle, it is necessary to ensure that at least at the final collision moment, the second obstacle avoidance position does not interfere with the second target edge point.

[0146] Coordination Figure 6 Taking the initial collision point Sin as an example, there may be two edge points C1 on the obstacle object with the same S-axis coordinate as the initial collision point and extreme L-axis coordinates. The method for determining the target edge point from these edge points can vary depending on the scenario. For example, in a scenario where lane changing is not considered, the distance between the edge point and the road boundary along the L-axis can be referenced, and the edge point with the larger distance can be selected as the first target edge point. Alternatively, in a scenario that considers lane changing, all of these edge points can be directly used as the first target edge point. The determination of the second target edge point is similar to that of the first target edge point and will not be elaborated upon here.

[0147] Coordination Figure 7 At the initial collision moment, the first obstacle avoidance position can be entirely located on the S-axis in the negative direction of the initial collision point, or the coordinates of the first obstacle avoidance position on the S-axis and L-axis may overlap with the initial collision point. At the final collision moment, the second obstacle avoidance position can also be entirely located on the S-axis in the positive direction of the final collision point, or the coordinates of the second obstacle avoidance position on the S-axis and L-axis may overlap with the final collision point.

[0148] In some embodiments, the width of the target vehicle is expanded using a preset vehicle body expansion coefficient, and the expanded vehicle width and the length of the target vehicle are used to approximate the projection of the target vehicle in the road coordinate system, thereby determining the first obstacle avoidance position and the second obstacle avoidance position.

[0149] Specifically, the L-axis coordinate range of the first obstacle avoidance position in the road coordinate system can be determined based on the first target edge point and the expanded vehicle width; and the S-axis coordinate range of the first obstacle avoidance position in the road coordinate system can be determined based on the first target edge point and the reference vehicle length of the target vehicle. Simultaneously, the L-axis coordinate range of the second obstacle avoidance position in the road coordinate system can be determined based on the second target edge point and the expanded vehicle width; and the S-axis coordinate range of the second obstacle avoidance position in the road coordinate system can be determined based on the second target edge point and the reference vehicle length of the target vehicle.

[0150] Coordination Figure 8 It is understandable that, in order to ensure that the first obstacle avoidance position does not interfere with the first target edge point at the initial collision moment, at least a portion of the target vehicle's body should preferably bypass the first target edge point, so that the S-axis coordinate of the first target edge point is within the S-axis coordinate range of the first obstacle avoidance position. Similarly, in order to ensure that the second obstacle avoidance position does not interfere with the second target edge point at the final collision moment, at least a portion of the target vehicle's body should also preferably bypass the second target edge point, so that the S-axis coordinate of the second target edge point is within the S-axis coordinate range of the second obstacle avoidance position.

[0151] Continue to cooperate with the participants Figure 8 Furthermore, in the road coordinate system, the L-axis coordinate of the first target edge point can be made equal to the extreme value of the L-axis coordinate interval of the first obstacle avoidance position, or not overlap with the L-axis coordinate interval of the first obstacle avoidance position. In this way, when the target vehicle is in the first obstacle avoidance position, the entire vehicle does not interfere with the first target edge point in the S-axis direction, increasing the probability of successful obstacle avoidance. Similarly, the L-axis coordinate of the second target edge point can also be made equal to the extreme value of the L-axis coordinate interval of the second obstacle avoidance position, or not overlap with the L-axis coordinate interval of the second obstacle avoidance position. In this way, when the target vehicle is in the second obstacle avoidance position, the entire vehicle also does not interfere with the second target edge point in the S-axis direction.

[0152] Coordination Figure 9 In one embodiment, in the road coordinate system, the S-axis coordinate range of the first obstacle avoidance position can be determined by identifying the center point of the first obstacle avoidance position and based on the center point of the first obstacle avoidance position and a reference vehicle length. The S-axis coordinate of the center point of the first obstacle avoidance position is equal to the S-axis coordinate of the edge point of the first target. Similarly, the S-axis coordinate range of the second obstacle avoidance position can be determined by identifying the center point of the second obstacle avoidance position and based on the center point of the second obstacle avoidance position and a reference vehicle length. The S-axis coordinate of the center point of the second obstacle avoidance position is equal to the S-axis coordinate of the edge point of the second target.

[0153] As an example, three circles O1, O2, and O3 connected end to end can be used to approximate the first obstacle avoidance position and the second obstacle avoidance position, wherein the diameter of each circle is equal to the expanded width of the target vehicle.

[0154] Taking the first obstacle avoidance position as an example, the center of circle O2 is the center point of the first obstacle avoidance position, and the S-axis coordinate of the center of circle O2 is equal to the S-axis coordinate of the edge point of the first target. The edge of circle O2 is tangent to the edge point of the first target. The line segment connecting the centers of circles O1, O2, and O3 is used as the reference vehicle length of the target vehicle. It can be seen that this reference vehicle length is twice the expanded vehicle width, and the reference vehicle length will change with the expansion of the vehicle width depending on the different vehicle expansion coefficients. The method for determining the second obstacle avoidance position can refer to the method for determining the first obstacle avoidance position, and will not be repeated here.

[0155] S14. Based on the planning start point, planning end point, and the first and second obstacle avoidance positions of the target vehicle, determine the obstacle avoidance reference path of the target vehicle in the local planning interval.

[0156] In one aspect, in the road coordinate system, the first obstacle avoidance position and the second obstacle avoidance position can be regarded as the projections of the target vehicle's outline (or expanded outline) at the initial collision time and the final collision time, respectively. Therefore, the target vehicle also has a corresponding heading at the first obstacle avoidance position and the second obstacle avoidance position. In this embodiment, the first obstacle avoidance position defines the target vehicle's first heading at the initial collision time, and the second obstacle avoidance position defines the target vehicle's second heading at the initial collision time.

[0157] Specifically, firstly, in the road coordinate system, a first reference line segment passing through the center point of the first obstacle avoidance position and extending along the first heading, and a second reference line segment passing through the center point of the second obstacle avoidance position and extending along the second heading, are determined, wherein the lengths of the first and second reference line segments are equal to the reference vehicle length. Secondly, the planning start point, the first reference line segment, the second reference line segment, and the planning end point are connected sequentially to obtain the obstacle avoidance reference path of the target vehicle in the local planning interval.

[0158] The first and second reference line segments can be regarded as vehicle length segments extending along the centerline of the target vehicle, and can be used to describe the path covered by the target vehicle in the first and second obstacle avoidance positions. Thus, by connecting the planning starting point, the first reference line segment, the second reference line segment, and the planning ending point with a straight line, an obstacle avoidance reference path that may avoid an obstacle can be described.

[0159] Coordination Figure 9The line connecting the centers of circles O1 and O3 at the initial and final collision times can be considered as the first and second reference line segments. The first and second reference line segments pass through the center point of circle O2, which is also the center point of the first obstacle avoidance position and the center point of the second obstacle avoidance position, respectively.

[0160] S15. Based on the obstacle avoidance reference path, plan the reference path for the target vehicle.

[0161] Coordination Figure 10 The obstacle avoidance reference path can be regarded as a "reference target" when planning the reference path of the target vehicle. For example, in the above embodiment, an initial reference path of the target vehicle is fitted with a fourth-order spline curve. The obstacle avoidance reference path can be divided into multiple sampling points according to the sampling interval (Δt) during path planning, and then participate in the replanning of the reference path of the target vehicle.

[0162] In the replanning of the reference path, it is desirable that the L-axis coordinate interval of the reference path within the local planning interval be as small as possible, which represents a smaller lateral displacement of the vehicle within the local planning interval. Therefore, in this embodiment, a sub-loss function is constructed based on the L-axis coordinate interval of the reference path within the local planning interval, and a comprehensive loss function is constructed based on this sub-loss function and the loss function of the preset path planning model; then, with the goal of minimizing the comprehensive loss function, the reference path of the target vehicle is planned without constraints based on the preset path planning model and the obstacle avoidance reference path.

[0163] In the above embodiments, the obstacle avoidance reference path was used separately during the replanning of the reference path. In other embodiments, it is desirable to minimize changes to the portion of the initial reference path after removing the local planning interval. Correspondingly, a fused reference path can be further constructed based on the obstacle avoidance reference path. This fused reference path is formed by connecting the portion of the initial reference path after removing the local planning interval with the obstacle avoidance reference path. Based on this fused reference path, the reference path of the target vehicle can then be planned. Similar to the previous embodiment, a comprehensive loss function can be constructed, and the reference path of the target vehicle can be planned unconstrainedly based on a preset path planning model and the fused reference path, with the goal of minimizing the comprehensive loss function.

[0164] As an example, taking the preset path planning model as a fourth-order spline curve model, it can have a loss function Loss1: the sum of squares of the fourth-order derivatives of the spline curve segments, used to ensure the smooth connection of each spline curve segment. The sub-loss function Loss2 can be the sum of squares of the L-axis coordinate interval of the reference path in the local planning interval. The comprehensive loss function Loss can be expressed as Loss1 + Loss2. Of course, in the construction of the comprehensive loss function Loss, Loss1 and Loss2 can also be weighted according to the needs of the actual scenario, and this application does not impose any restrictions on this.

[0165] After replanning the reference path, this application further proposes to perform curvature detection and collision detection operations on the reference path to confirm whether the replanned reference path meets the corresponding requirements. The curvature detection and collision detection are explained separately below.

[0166] ① Curvature detection

[0167] One purpose of curvature detection is to confirm whether a convex hull exists on the reference path. The presence of a convex hull may increase the risk for the target vehicle when traveling along the reference path due to excessive curvature. Therefore, in this embodiment, it is confirmed whether the maximum curvature on the reference path exceeds a preset curvature threshold; if so, the length of the local planning interval is expanded to replan the reference path for the target vehicle; if not, the curvature detection is considered successful.

[0168] One reason for excessive curvature is the limited length of the local planning interval, which refers to the length of the local planning interval along the S-axis of the road coordinate system. In a specific embodiment, a preset obstacle avoidance adjustment time can be increased (e.g., by 0.2 seconds each time) to adjust the S-axis coordinates of the planning start point and / or planning end point of the local planning interval. Based on the expanded local planning interval, the replanned reference path can reduce the possibility of convex hull. Furthermore, in this embodiment, the above curvature detection steps can be iteratively performed on the replanned reference path until the maximum curvature on the reference path meets the expectation.

[0169] ② Collision Detection

[0170] Since the planning of the reference path in this embodiment can be unconstrained, it is also possible to detect whether there is a collision point between the reference path and the obstacle object; if so, the L-axis coordinate interval or the L-axis and S-axis coordinate interval of the first obstacle avoidance position and the second obstacle avoidance position in the road coordinate system is expanded to replan the reference path of the target vehicle; if not, the collision detection is deemed to have passed.

[0171] One reason for collision detection failure is that the lateral dimensions of the first and second obstacle avoidance positions are too small, causing the obstacle avoidance reference path to be too close to the obstacle object in the L-axis direction; or, the lateral and longitudinal dimensions of both the first and second obstacle avoidance positions are small. In a specific embodiment, a preset vehicle body expansion coefficient can be increased (e.g., by 0.2 each time), thereby expanding the outline of the first and second obstacle avoidance positions that approach the target vehicle projection. Based on the expanded first and second obstacle avoidance positions, the replanned reference path can reduce the possibility of collision with the obstacle object. Similarly, in this embodiment, the above collision detection steps can be iteratively executed on the replanned reference path until the reference path will not collide with the obstacle object.

[0172] In different embodiments of the present application, the execution order of the above curvature detection and collision detection can be adjusted according to the requirements of scenarios. For example, curvature detection may be performed on a reference path first until a reference path that passes curvature detection is obtained; then, collision detection is performed on the reference path that has passed curvature detection until a reference path that passes collision detection is obtained, which is used as the final reference path of the target vehicle at this time. For another example, collision detection may be performed on a reference path first until a reference path that passes collision detection is obtained; then, curvature detection is performed on the reference path that has passed collision detection until a reference path that passes curvature detection is obtained, which is used as the final reference path of the target vehicle at this time.

[0173] Reference Figure 11 , an embodiment of the path planning device of the present application is introduced. In this embodiment, the path planning device includes an acquisition module 21, a first determination module 22, a second determination module 23, a third determination module 24, and a planning module 25.

[0174] The acquisition module is configured to acquire an initial reference path of a target vehicle, and an initial collision point and a final collision point with an obstacle object; the first determination module is configured to determine a local planning interval of the target vehicle based on the initial collision point and the final collision point, wherein the local planning interval extends along the S-axis direction of a road coordinate system, and the local planning interval intersects the initial reference path at a planning start point and a planning end point; the second determination module is configured to determine a first obstacle avoidance position of the target vehicle at an initial collision moment based on the initial collision point, and determine a second obstacle avoidance position of the target vehicle at a final collision moment based on the final collision point; the third determination module is configured to determine an obstacle avoidance reference path of the target vehicle in the local planning interval based on the planning start point, the planning end point, as well as the first obstacle avoidance position and the second obstacle avoidance position of the target vehicle; the planning module is configured to plan a reference path for the target vehicle based on the obstacle avoidance reference path.

[0175] In one embodiment, the first determination module is specifically configured to divide an obstacle avoidance adjustment duration into a first adjustment duration and a second adjustment duration based on a preset rule; determine a first adjustment value based on the first adjustment duration and the vehicle speed of the target vehicle, and determine a second adjustment value based on the second adjustment duration and the vehicle speed of the target vehicle; on the initial reference path, reversely adjust by the first adjustment value from the initial collision point to determine the planning start point of the local planning interval, and forward adjust by the second adjustment value from the final collision point to determine the planning end point of the local planning interval.

[0176] In one embodiment, the obstacle object includes a first obstacle object and a second obstacle object; the first determining module is specifically used to determine a first sub-local planning interval of the target vehicle based on the initial collision point and the final collision point of the initial reference path and the first obstacle object, and to determine a second sub-local planning interval of the target vehicle based on the initial collision point and the final collision point of the initial reference path and the second obstacle object, wherein the first sub-local planning interval and the second sub-local planning interval extend along the S-axis direction of the road coordinate system; determine whether the first sub-local planning interval and the second sub-local planning interval include an overlapping interval; if so, merge the first sub-local planning interval and the second sub-local planning interval into the local planning interval.

[0177] In one embodiment, the second determining module is specifically used to determine a first target edge point corresponding to the obstacle object and the initial collision point, wherein, in the road coordinate system, the L-axis coordinate of the first target edge point is an extreme value and equal to the S-axis coordinate of the initial collision point; based on the first target edge point and a preset vehicle body expansion coefficient, a first obstacle avoidance position of the target vehicle at the time of the initial collision is determined; and / or, the second determining module is specifically used to determine a second target edge point corresponding to the obstacle object and the final collision point, wherein, in the road coordinate system, the L-axis coordinate of the second target edge point is an extreme value and equal to the S-axis coordinate of the final collision point; based on the second target edge point and a preset vehicle body expansion coefficient, a second obstacle avoidance position of the target vehicle at the time of the final collision is determined.

[0178] In one embodiment, the second determining module is specifically used to determine the L-axis coordinate interval of the first obstacle avoidance position in the road coordinate system based on the first target edge point and the expanded vehicle width, wherein the expanded vehicle width is determined by the vehicle width expansion of the target vehicle using the preset vehicle body expansion coefficient; determine the S-axis coordinate interval of the first obstacle avoidance position in the road coordinate system based on the first target edge point and the reference vehicle length of the target vehicle, wherein the S-axis coordinate of the first target edge point is located within the S-axis coordinate interval of the first obstacle avoidance position; and / or, the second determining module is specifically used to determine the L-axis coordinate interval of the second obstacle avoidance position in the road coordinate system based on the second target edge point and the expanded vehicle width, wherein the expanded vehicle width is determined by the vehicle width expansion of the target vehicle using the preset vehicle body expansion coefficient; determine the S-axis coordinate interval of the second obstacle avoidance position in the road coordinate system based on the second target edge point and the reference vehicle length of the target vehicle, wherein the S-axis coordinate of the second target edge point is located within the S-axis coordinate interval of the second obstacle avoidance position.

[0179] In one embodiment, the second determining module is specifically used to determine the center point of the first obstacle avoidance position in the road coordinate system, wherein the S-axis coordinate of the center point of the first obstacle avoidance position is equal to the S-axis coordinate of the edge point of the first target; and to determine the S-axis coordinate range of the first obstacle avoidance position based on the center point of the first obstacle avoidance position and the reference vehicle length; and / or, the second determining module is specifically used to determine the center point of the second obstacle avoidance position in the road coordinate system, wherein the S-axis coordinate of the center point of the second obstacle avoidance position is equal to the S-axis coordinate of the edge point of the second target; and to determine the S-axis coordinate range of the second obstacle avoidance position based on the center point of the second obstacle avoidance position and the reference vehicle length.

[0180] In one embodiment, the first obstacle avoidance position is used to define a first heading of the target vehicle at the initial collision moment, and the second obstacle avoidance position is used to define a second heading of the target vehicle at the final collision moment; the third determining module is specifically used to determine, in the road coordinate system, a first reference line segment passing through the center point of the first obstacle avoidance position and extending along the first heading, and a second reference line segment passing through the center point of the second obstacle avoidance position and extending along the second heading, wherein the lengths of the first reference line segment and the second reference line segment are equal to the reference vehicle length; the planning starting point, the first reference line segment, the second reference line segment, and the planning ending point are connected sequentially to obtain the obstacle avoidance reference path of the target vehicle in the local planning interval.

[0181] In one embodiment, the planning module is specifically used to construct a fused reference path based on the obstacle avoidance reference path, wherein the fused reference path is formed by connecting the initial reference path after removing the portion within the local planning interval and the obstacle avoidance reference path; and to plan a reference path for the target vehicle based on the fused reference path.

[0182] In one embodiment, the planning module is specifically used to construct a sub-loss function in the road coordinate system based on the L-axis coordinate interval of the obstacle avoidance reference path in the local planning interval; construct a comprehensive loss function based on the sub-loss function and the loss function of the preset path planning model; and plan the reference path of the target vehicle without constraints based on the preset path planning model and the fused reference path with the goal of minimizing the comprehensive loss function.

[0183] In one embodiment, the planning module is further configured to confirm whether the maximum curvature on the reference path exceeds a preset curvature threshold; if so, to expand the interval length of the local planning interval in order to replan the reference path of the target vehicle.

[0184] In one embodiment, the planning module is further configured to detect whether there is a collision point between the reference path and the obstacle object; if so, expand the L-axis coordinate interval or the L-axis and S-axis coordinate interval of the first obstacle avoidance position and the second obstacle avoidance position in the road coordinate system to replan the reference path of the target vehicle.

[0185] As referred above Figures 1 to 10 The path planning method according to embodiments of this specification has been described. The details mentioned in the above description of the method embodiments also apply to the path planning apparatus of embodiments of this specification. The above-described path planning apparatus can be implemented in hardware, software, or a combination of hardware and software.

[0186] Figure 12 A hardware structure diagram of an unmanned vehicle according to an embodiment of this specification is shown. Figure 12 As shown, the unmanned vehicle 30 may include at least one processor 31, a memory 32 (e.g., non-volatile memory), a RAM 33, and a communication interface 34, and the at least one processor 31, memory 32, RAM 33, and communication interface 34 are connected together via an internal bus 35. At least one processor 31 executes at least one computer-readable instruction stored or encoded in the memory 32.

[0187] It should be understood that the computer-executable instructions stored in memory 32, when executed, cause at least one processor 31 to perform the above-described combinations in the various embodiments of this specification. Figures 1 to 10 The description includes various operations and functions.

[0188] In the embodiments of this specification, the unmanned vehicle 30 can be configured with a functional terminal to carry the above-mentioned hardware structure. The terminal may include, but is not limited to: personal computer, server computer, workstation, desktop computer, laptop computer, notebook computer, mobile electronic device, smartphone, tablet computer, cellular phone, personal digital assistant (PDA), handheld device, messaging device, wearable electronic device, consumer electronic device, etc.

[0189] According to one embodiment, a program product, such as a machine-readable medium, is provided. The machine-readable medium may have instructions (i.e., the elements implemented in software as described above), which, when executed by a machine, cause the machine to perform the above-described combinations of the various embodiments of this specification. Figures 1 to 10 The various operations and functions described. Specifically, a system or apparatus equipped with a readable storage medium storing software program code that implements the functions of any of the embodiments described above, and enabling the computer or processor of the system or apparatus to read and execute the instructions stored in the readable storage medium.

[0190] In this case, the program code read from the readable medium itself can perform the functions of any of the above embodiments, and therefore the machine-readable code and the readable storage medium storing the machine-readable code constitute a part of this specification.

[0191] Examples of readable storage media include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD-RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, program code can be downloaded from a server computer or the cloud via a communication network.

[0192] Those skilled in the art will understand that the various embodiments disclosed above can be modified and varied without departing from the spirit of the invention. Therefore, the scope of protection of this specification should be defined by the appended claims.

[0193] It should be noted that not all steps and units in the above process and system structure diagrams are mandatory; some steps or units can be omitted according to actual needs. The execution order of each step is not fixed and can be determined as needed. The device structure described in the above embodiments can be a physical structure or a logical structure. That is, some units may be implemented by the same physical client, or some units may be implemented by multiple physical clients, or they may be jointly implemented by certain components in multiple independent devices.

[0194] In the above embodiments, the hardware units or modules can be implemented mechanically or electrically. For example, a hardware unit, module, or processor may include permanent dedicated circuitry or logic (such as a dedicated processor, FPGA, or ASIC) to perform the corresponding operation. The hardware unit or processor may also include programmable logic or circuitry (such as a general-purpose processor or other programmable processor), which can be temporarily configured by software to perform the corresponding operation. The specific implementation method (mechanical, dedicated permanent circuitry, or temporarily configured circuitry) can be determined based on cost and time considerations.

[0195] The specific embodiments described above with reference to the accompanying drawings are exemplary embodiments, but do not represent all embodiments that can be implemented or fall within the scope of the claims. The term "exemplary" as used throughout this specification means "serving as an example, instance, or illustration" and does not imply that it is "preferred" or "advantageous" compared to other embodiments. Specific details are included to provide an understanding of the described techniques. However, these techniques can be practiced without these specific details. In some instances, well-known structures and apparatuses are shown in block diagram form to avoid obscuring the concepts of the described embodiments.

[0196] The foregoing description of this disclosure is provided to enable any person skilled in the art to implement or use this disclosure. Various modifications to this disclosure will be apparent to those skilled in the art, and the general principles applicable herein can be applied to other variations without departing from the scope of this disclosure. Therefore, this disclosure is not limited to the examples and designs described herein, but is consistent with the widest scope of the principles and novel features disclosed herein.

Claims

1. A path planning method, characterized in that, The method includes: Obtain the initial reference path of the target vehicle and the initial and final collision points of the obstacle objects; Based on the initial collision point and the final collision point, a local planning interval for the target vehicle is determined, wherein the local planning interval extends along the S-axis direction of the road coordinate system, and the local planning interval intersects with the initial reference path at the planning start point and the planning end point. Determining the first obstacle avoidance position of the target vehicle at the initial collision moment based on the initial collision point specifically includes determining the first target edge point corresponding to the obstacle object and the initial collision point, wherein, in the road coordinate system, the L-axis coordinate of the first target edge point takes an extreme value and is equal to the S-axis coordinate of the initial collision point; and determining the first obstacle avoidance position of the target vehicle at the initial collision moment based on the first target edge point and a preset vehicle body expansion coefficient. Determining the second obstacle avoidance position of the target vehicle at the moment of the final collision based on the terminal collision point specifically includes determining the second target edge point corresponding to the obstacle object and the terminal collision point, wherein, in the road coordinate system, the L-axis coordinate of the second target edge point takes an extreme value and is equal to the S-axis coordinate of the terminal collision point; and determining the second obstacle avoidance position of the target vehicle at the moment of the final collision based on the second target edge point and a preset vehicle body expansion coefficient. Based on the planning start point, planning end point, and the first and second obstacle avoidance positions of the target vehicle, determine the obstacle avoidance reference path of the target vehicle in the local planning interval; Based on the obstacle avoidance reference path, a reference path for the target vehicle is planned.

2. The path planning method according to claim 1, characterized in that, Based on the initial collision point and the final collision point, the local planning interval of the target vehicle is determined, specifically including: Based on preset rules, the obstacle avoidance adjustment time is divided into a first adjustment time and a second adjustment time; A first adjustment value is determined based on the first adjustment duration and the speed of the target vehicle, and a second adjustment value is determined based on the second adjustment duration and the speed of the target vehicle. On the initial reference path, the planning start point of the local planning interval is determined by adjusting the first adjustment value in reverse from the initial collision point, and the planning end point of the local planning interval is determined by adjusting the second adjustment value in forward from the final collision point.

3. The path planning method according to claim 1, characterized in that, The obstacle object includes a first obstacle object and a second obstacle object; Based on the initial collision point and the final collision point, the local planning interval of the target vehicle is determined, specifically including: Based on the initial collision point and final collision point of the initial reference path and the first obstacle object, a first sub-local planning interval of the target vehicle is determined, and based on the initial collision point and final collision point of the initial reference path and the second obstacle object, a second sub-local planning interval of the target vehicle is determined, wherein the first sub-local planning interval and the second sub-local planning interval extend along the S-axis direction of the road coordinate system. Determine whether the first sub-local planning interval and the second sub-local planning interval include overlapping intervals; if so, The first sub-local planning interval and the second sub-local planning interval are merged into the local planning interval.

4. The path planning method according to claim 1, characterized in that, Based on the first target edge point and a preset vehicle body expansion coefficient, the first obstacle avoidance position of the target vehicle at the initial collision moment is determined, specifically including: Based on the first target edge point and the expanded vehicle width, the L-axis coordinate range of the first obstacle avoidance position in the road coordinate system is determined, wherein the expanded vehicle width is determined by the expansion of the target vehicle width by the preset vehicle body expansion coefficient. The S-axis coordinate range of the first obstacle avoidance position in the road coordinate system is determined based on the first target edge point and the reference vehicle length of the target vehicle, wherein the S-axis coordinate of the first target edge point is located within the S-axis coordinate range of the first obstacle avoidance position. And / or, Based on the second target edge point and a preset vehicle body expansion coefficient, the second obstacle avoidance position of the target vehicle at the moment of final collision is determined, specifically including: Based on the second target edge point and the expanded vehicle width, the L-axis coordinate range of the second obstacle avoidance position in the road coordinate system is determined, wherein the expanded vehicle width is determined by the expansion of the target vehicle width by the preset vehicle body expansion coefficient; The S-axis coordinate range of the second obstacle avoidance position in the road coordinate system is determined based on the second target edge point and the reference vehicle length of the target vehicle, wherein the S-axis coordinate of the second target edge point is located within the S-axis coordinate range of the second obstacle avoidance position.

5. The path planning method according to claim 4, characterized in that, The S-axis coordinate range of the first obstacle avoidance position in the road coordinate system is determined based on the first target edge point and the reference vehicle length of the target vehicle, specifically including: In the road coordinate system, the center point of the first obstacle avoidance position is determined, wherein the S-axis coordinate of the center point of the first obstacle avoidance position is equal to the S-axis coordinate of the edge point of the first target. Based on the center point of the first obstacle avoidance position and the reference vehicle length, determine the S-axis coordinate range of the first obstacle avoidance position; And / or, The S-axis coordinate range of the second obstacle avoidance position in the road coordinate system is determined based on the second target edge point and the reference vehicle length of the target vehicle, specifically including: In the road coordinate system, the center point of the second obstacle avoidance position is determined, wherein the S-axis coordinate of the center point of the second obstacle avoidance position is equal to the S-axis coordinate of the edge point of the second target. Based on the center point of the second obstacle avoidance position and the reference vehicle length, the S-axis coordinate range of the second obstacle avoidance position is determined.

6. The path planning method according to claim 4, characterized in that, The reference vehicle length is twice the width of the expanded vehicle.

7. The path planning method according to claim 4, characterized in that, The first obstacle avoidance position is used to define the first heading of the target vehicle at the initial collision time, and the second obstacle avoidance position is used to define the second heading of the target vehicle at the final collision time; Based on the planning start point, planning end point, and the first and second obstacle avoidance positions of the target vehicle, an obstacle avoidance reference path for the target vehicle within the local planning interval is determined, specifically including: In the road coordinate system, a first reference line segment is determined that passes through the center point of the first obstacle avoidance position and extends along the first heading, and a second reference line segment is determined that passes through the center point of the second obstacle avoidance position and extends along the second heading, wherein the lengths of the first reference line segment and the second reference line segment are equal to the reference vehicle length. By sequentially connecting the planning starting point, the first reference line segment, the second reference line segment, and the planning ending point, the obstacle avoidance reference path of the target vehicle in the local planning interval is obtained.

8. The path planning method according to claim 1, characterized in that, Based on the obstacle avoidance reference path, a reference path for the target vehicle is planned, specifically including: Based on the obstacle avoidance reference path, a fusion reference path is constructed, wherein the fusion reference path is formed by connecting the initial reference path after removing the portion within the local planning interval and the obstacle avoidance reference path; Based on the fused reference path, a reference path for the target vehicle is planned.

9. The path planning method according to claim 8, characterized in that, Based on the fused reference path, a reference path for the target vehicle is planned, specifically including: In the road coordinate system, a sub-loss function is constructed based on the L-axis coordinate interval of the obstacle avoidance reference path in the local planning interval; Based on the sub-loss function and the loss function of the preset path planning model, a comprehensive loss function is constructed; With the goal of minimizing the comprehensive loss function, the reference path of the target vehicle is planned without constraints based on the preset path planning model and the fused reference path.

10. The path planning method according to claim 1, characterized in that, The method further includes: Confirm whether the maximum curvature on the reference path exceeds a preset curvature threshold; if so, The length of the local planning interval is expanded to replan the reference path of the target vehicle.

11. The path planning method according to claim 1, characterized in that, The method further includes: Detect whether there is a collision point between the reference path and the obstacle object; if so, Expand the L-axis coordinate range, or the L-axis and S-axis coordinate range, of the first and second obstacle avoidance positions in the road coordinate system to replan the reference path of the target vehicle.

12. A path planning device, characterized in that, include: The acquisition module is used to acquire the initial reference path of the target vehicle and the initial and final collision points of the obstacle objects; The first determining module is used to determine the local planning interval of the target vehicle based on the initial collision point and the final collision point, wherein the local planning interval extends along the S-axis direction of the road coordinate system, and the local planning interval intersects the initial reference path at the planning start point and the planning end point. The second determining module is used to determine the first obstacle avoidance position of the target vehicle at the initial collision moment based on the initial collision point. Specifically, it is used to determine the first target edge point corresponding to the obstacle object and the initial collision point, wherein, in the road coordinate system, the L-axis coordinate of the first target edge point takes an extreme value and is equal to the S-axis coordinate of the initial collision point; based on the first target edge point and a preset vehicle body expansion coefficient, it determines the first obstacle avoidance position of the target vehicle at the initial collision moment, and The method is used to determine the second obstacle avoidance position of the target vehicle at the moment of final collision based on the final collision point. Specifically, it is used to determine the second target edge point corresponding to the obstacle object and the final collision point, wherein, in the road coordinate system, the L-axis coordinate of the second target edge point takes an extreme value and is equal to the S-axis coordinate of the final collision point; the second obstacle avoidance position of the target vehicle at the moment of final collision is determined based on the second target edge point and a preset vehicle body expansion coefficient. The third determining module is used to determine the obstacle avoidance reference path of the target vehicle in the local planning interval based on the planning start point, planning end point, and the first obstacle avoidance position and the second obstacle avoidance position of the target vehicle; The planning module is used to plan a reference path for the target vehicle based on the obstacle avoidance reference path.

13. An unmanned vehicle, comprising: At least one processor; as well as A memory that stores instructions that, when executed by the at least one processor, cause the at least one processor to perform the path planning method as described in any one of claims 1 to 11.

14. A machine-readable storage medium storing executable instructions that, when executed, cause the machine to perform the path planning method as described in any one of claims 1 to 11.

Images