Detour trajectory planning method, device, equipment and storage medium

Abstract

The present application provides a detour trajectory planning method, device, equipment, and storage medium, relating to the field of intelligent driving technology. The detour trajectory planning method includes: constructing a longitudinal feasible space map corresponding to the target vehicle's detour around obstacles within a preset time period in the future, and obtaining a detour reference speed for the target vehicle based on the longitudinal feasible space map; sampling and evaluating the feasible space corresponding to the detour behavior stage of the target vehicle based on the detour reference speed to determine a driving reference path; and determining a target detour trajectory for the target vehicle within a preset time period in the future based on the longitudinal feasible space map and the driving reference path. The present application can greatly reduce driving risks, improve driving efficiency, and make driving more flexible and comfortable.

Description

Technical Field

[0001] The present application relates to the field of intelligent driving technology, and in particular to a detour trajectory planning method, device, equipment and storage medium. Background Art

[0002] With the development of autonomous driving technology, autonomous vehicles are gradually being developed and applied. When an autonomous vehicle is driving, a planned driving trajectory is provided to the autonomous vehicle so that the autonomous vehicle can automatically drive according to the planned driving trajectory.

[0003] Currently, when an autonomous vehicle senses an obstacle partially entering its lane, it typically plans its trajectory to proactively overtake the obstacle by changing lanes. However, this lane-changing approach increases driving risks. Summary of the Invention

[0004] The present application provides a detour trajectory planning method, device, equipment and storage medium to solve the problem of increased driving risks caused by actively overtaking obstacles by changing lanes.

[0005] In a first aspect, the present application provides a method for planning a detour trajectory, comprising:

[0006] Construct a longitudinal feasible space diagram corresponding to the target vehicle's detour around obstacles within a preset future time. The longitudinal feasible space diagram is used to represent the longitudinal feasible range and longitudinal speed constraint information corresponding to discrete moments in the target vehicle's future preset time.

[0007] Based on the longitudinal feasible space diagram, the detour reference speed of the target vehicle is obtained;

[0008] Based on the detour reference speed, the feasible space corresponding to the target vehicle's detour behavior phase is sampled and evaluated to determine the driving reference path. The detour behavior phase includes the pre-detour preparation phase, the detour process phase, and the phase of returning to the lane center after the detour is completed. The detour behavior phase is determined based on the longitudinal feasible space window corresponding to the target vehicle's detour around the obstacle within a preset time period in the future.

[0009] Based on the longitudinal feasible space map and the driving reference path, the target detour trajectory of the target vehicle within a preset time period in the future is determined.

[0010] Optionally, based on the detour reference speed, the feasible space corresponding to the detour behavior stage of the target vehicle is sampled and evaluated to determine the driving reference path, including: sampling the feasible space corresponding to the detour behavior stage of the target vehicle to obtain corresponding node information; determining the trajectory corresponding to the target edge based on the detour reference speed and edge information, the edge information is the edge information obtained by connecting the nodes included in two adjacent detour behavior stages; obtaining the cost value of the trajectory corresponding to the target edge according to a first preset cost function, the first preset cost function is used to express the quantitative relationship between the cost value and the trajectory difference, speed, collision risk and maximum yaw angular velocity of the corresponding trajectory; determining the driving reference path based on the cost value.

[0011] Optionally, the feasible space corresponding to the bypass behavior stage of the target vehicle is sampled to obtain corresponding node information, including: sampling the longitudinal feasible space of each bypass behavior stage at equal intervals to obtain the corresponding discrete longitudinal feasible distance; based on the discrete longitudinal feasible distance, lane width and position information of the obstacle to be bypassed, sampling the lateral feasible space of the target vehicle at equal intervals to obtain corresponding sampling points; and obtaining corresponding node information based on the sampling points.

[0012] Optionally, determining the trajectory corresponding to the target edge according to the detour reference speed and the edge information includes: discretizing and interpolating the longitudinal feasible distance range corresponding to the edge according to the detour reference speed to determine the trajectory corresponding to the target edge.

[0013] Optionally, a driving reference path is determined based on the cost value, including: taking the initial position of the target vehicle as the root node, connecting each edge included in the detour behavior stage from the root node in sequence to obtain a node connection sequence; determining the node connection sequence with the minimum cost value addition; and determining the driving reference path based on the edge corresponding to the node connection sequence with the minimum cost value addition.

[0014] Optionally, a longitudinal feasible space diagram corresponding to the target vehicle bypassing an obstacle within a preset time period in the future is constructed, including: determining the target longitudinal feasible minimum distance, target longitudinal feasible maximum distance, minimum target longitudinal speed, and maximum target longitudinal speed corresponding to the target vehicle bypassing the obstacle at discrete moments within the preset time period in the future; and constructing the longitudinal feasible space diagram based on the target longitudinal feasible minimum distance, target longitudinal feasible maximum distance, minimum target longitudinal speed, and maximum target longitudinal speed.

[0015] Optionally, determining the target longitudinal feasible minimum distance, target longitudinal feasible maximum distance, minimum target longitudinal speed, and maximum target longitudinal speed corresponding to the target vehicle's detour around obstacles at discrete moments in a preset future time period, including: determining the target vehicle's initial longitudinal feasible maximum distance, initial longitudinal feasible minimum distance, minimum initial longitudinal speed, and maximum initial longitudinal speed; for the obstacles to be detoured at discrete moments in the preset future time period, performing the following operations until all obstacles are traversed: determining the target longitudinal feasible minimum distance of the target vehicle based on the longitudinal space occupied by the target obstacle and the initial longitudinal feasible minimum distance, and determining the minimum value of the target longitudinal speed based on the speed of the target obstacle, determining the target longitudinal feasible minimum distance as the new initial longitudinal feasible minimum distance, and determining the minimum value of the target longitudinal speed as the new minimum value of the initial longitudinal speed.

[0016] Optionally, based on the longitudinal feasible space diagram, a detour reference speed of the target vehicle is obtained, including: based on the longitudinal feasible space diagram, according to the maximum acceleration capability, maximum deceleration capability, the curvature corresponding to the reference line of the path within the longitudinal feasible range, and the correspondence between the preset curvature and the speed constraint, obtaining updated longitudinal feasible range and longitudinal speed constraint information; based on the updated longitudinal feasible range and longitudinal speed constraint information, obtaining an updated longitudinal feasible space diagram; based on the updated longitudinal feasible space diagram and a second preset cost function, obtaining a first target position of the target vehicle corresponding to a discrete moment, the second preset cost function being used to express the quantitative relationship between the cost value and the speed, acceleration and jerk of the sampling point corresponding to the discrete moment; fitting processing is performed on the first target position to obtain a reference speed of the target vehicle.

[0017] Optionally, based on the longitudinal feasible space map and the driving reference path, the target bypass trajectory of the target vehicle within a preset time period in the future is determined, including: performing speed sampling based on the longitudinal feasible space map to obtain sampled alternative speeds; generating alternative bypass trajectories based on the sampled alternative speeds and the driving reference path; determining the target bypass trajectory based on the alternative bypass trajectory and a third preset cost function, the third preset cost function being used to express the quantitative relationship between the cost value and the collision risk, maximum acceleration, maximum jerk and maximum yaw angular velocity of the corresponding alternative trajectory.

[0018] Optionally, before speed sampling is performed based on the longitudinal feasible space graph, the detour trajectory planning method further includes: updating longitudinal speed constraint information of the longitudinal feasible space graph according to the driving reference path, the position information and the type attribute of the obstacle.

[0019] Optionally, before constructing a longitudinal feasible space map corresponding to the target vehicle's detour around obstacles within a preset time period in the future, the detour trajectory planning method also includes: clustering the obstacles of the target vehicle within the preset time period in the future to obtain corresponding clustering processing results; performing detour feasibility analysis based on the clustering processing results; and determining that the target vehicle can detour around obstacles within the preset time period in the future.

[0020] In a second aspect, the present application provides a detour trajectory planning device, comprising:

[0021] A construction module is used to construct a longitudinal feasible space diagram corresponding to the target vehicle's detour around obstacles within a preset time period in the future. The longitudinal feasible space diagram is used to represent the longitudinal feasible range and longitudinal speed constraint information corresponding to discrete moments in the target vehicle within the preset time period in the future;

[0022] An acquisition module, used for obtaining a detour reference speed of the target vehicle based on the longitudinal feasible space graph;

[0023] A processing module is configured to sample and evaluate the feasible space corresponding to the target vehicle's detour behavior phase based on the detour reference speed to determine a driving reference path. The detour behavior phase includes a pre-detour preparation phase, a detour process phase, and a post-detour return to the lane center phase. The detour behavior phase is determined based on a longitudinal feasible space window corresponding to the target vehicle's detour around the obstacle within a preset future time period.

[0024] The first determination module is used to determine a target detour trajectory of the target vehicle within a preset time period in the future based on the longitudinal feasible space map and the driving reference path.

[0025] Optionally, the processing module is specifically used to: sample the feasible space corresponding to the detour behavior stage of the target vehicle to obtain corresponding node information; determine the trajectory corresponding to the target edge based on the detour reference speed and edge information, the edge information is the edge information obtained by connecting the nodes included in two adjacent detour behavior stages; obtain the cost value of the trajectory corresponding to the target edge according to a first preset cost function, the first preset cost function is used to express the quantitative relationship between the cost value and the trajectory difference, speed, collision risk and maximum yaw angular velocity of the corresponding trajectory; determine the driving reference path based on the cost value.

[0026] Optionally, when the processing module is used to sample the feasible space corresponding to the bypass behavior stage of the target vehicle and obtain the corresponding node information, it is specifically used to: sample the longitudinal feasible space of each bypass behavior stage at equal intervals to obtain the corresponding discrete longitudinal feasible distance; based on the discrete longitudinal feasible distance, lane width and position information of the obstacle to be bypassed, sample the lateral feasible space of the target vehicle at equal intervals to obtain the corresponding sampling points; and obtain the corresponding node information based on the sampling points.

[0027] Optionally, when the processing module is used to determine the trajectory corresponding to the target edge based on the detour reference speed and edge information, it is specifically used to: discretize and interpolate the longitudinal feasible distance range corresponding to the edge based on the detour reference speed to determine the trajectory corresponding to the target edge.

[0028] Optionally, when the processing module is used to determine the driving reference path based on the cost value, it is specifically used to: take the initial position of the target vehicle as the root node, connect each edge included in the detour behavior stage from the root node in sequence to obtain a node connection sequence; determine the node connection sequence with the minimum cost value addition; and determine the driving reference path based on the edge corresponding to the node connection sequence with the minimum cost value addition.

[0029] Optionally, the construction module is specifically used to: determine the target longitudinal feasible minimum distance, target longitudinal feasible maximum distance, minimum target longitudinal speed and maximum target longitudinal speed corresponding to the target vehicle's detour around obstacles at discrete moments within a preset time period in the future; and construct a longitudinal feasible space diagram based on the target longitudinal feasible minimum distance, target longitudinal feasible maximum distance, minimum target longitudinal speed and maximum target longitudinal speed.

[0030] Optionally, when the construction module is used to determine the target feasible minimum longitudinal distance, target feasible maximum distance, minimum target longitudinal speed, and maximum target longitudinal speed corresponding to the target vehicle circumventing obstacles at discrete moments in the future preset time period, it is specifically used to: determine the initial feasible maximum longitudinal distance, initial feasible minimum longitudinal distance, minimum initial longitudinal speed, and maximum initial longitudinal speed of the target vehicle; for the obstacles to be circumvented at discrete moments in the future preset time period, perform the following operations until all obstacles are traversed: determine the target feasible minimum longitudinal distance of the target vehicle based on the longitudinal space occupied by the target obstacle and the initial feasible minimum longitudinal distance, and determine the minimum value of the target longitudinal speed based on the speed of the target obstacle, determine the target feasible minimum distance as the new initial feasible minimum longitudinal distance, and determine the minimum value of the target longitudinal speed as the new minimum value of the initial longitudinal speed.

[0031] Optionally, the acquisition module is specifically used to: based on the longitudinal feasible space diagram, obtain the updated longitudinal feasible range and longitudinal speed constraint information according to the maximum acceleration capability, maximum deceleration capability, the curvature corresponding to the reference line of the path within the longitudinal feasible range, and the correspondence between the preset curvature and the speed constraint; obtain the updated longitudinal feasible space diagram according to the updated longitudinal feasible range and longitudinal speed constraint information; based on the updated longitudinal feasible space diagram and the second preset cost function, obtain the first target position of the target vehicle corresponding to the discrete moment, the second preset cost function being used to express the quantitative relationship between the cost value and the speed, acceleration and jerk of the sampling point corresponding to the discrete moment; perform fitting processing on the first target position to obtain the reference speed of the target vehicle.

[0032] Optionally, the first determination module is specifically used to: perform speed sampling based on the longitudinal feasible space diagram to obtain sampled alternative speeds; generate alternative detour trajectories based on the sampled alternative speeds and the driving reference path; determine the target detour trajectory based on the alternative detour trajectory and a third preset cost function, and the third preset cost function is used to express the quantitative relationship between the cost value and the collision risk, maximum acceleration, maximum jerk and maximum yaw angular velocity of the corresponding alternative trajectory.

[0033] Optionally, the detour trajectory planning device also includes an updating module for updating the longitudinal speed constraint information of the longitudinal feasible space graph according to the driving reference path, the position information and type attributes of the obstacle before the first determination module performs speed sampling based on the longitudinal feasible space graph.

[0034] Optionally, the detour trajectory planning device also includes a second determination module, which is used to cluster the obstacles of the target vehicle within a future preset time period before the construction module constructs the longitudinal feasible space map corresponding to the target vehicle's detour around obstacles within a future preset time period to obtain corresponding clustering processing results; perform detour feasibility analysis based on the clustering processing results; and determine that the target vehicle can detour around obstacles within a future preset time period.

[0035] In a third aspect, the present application provides an electronic device, comprising: a processor, and a memory communicatively connected to the processor;

[0036] Memory stores computer-executable instructions;

[0037] The processor executes the computer-executable instructions stored in the memory to implement the bypass trajectory planning method as described in the first aspect of the present application.

[0038] In a fourth aspect, the present application provides a computer-readable storage medium, in which computer program instructions are stored. When the computer program instructions are executed by a processor, the bypass trajectory planning method described in the first aspect of the present application is implemented.

[0039] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the bypass trajectory planning method as described in the first aspect of the present application.

[0040] The detour trajectory planning method, device, equipment and storage medium provided by the present application construct a longitudinal feasible space map corresponding to the target vehicle's detour around obstacles within a preset time period in the future, and obtain the target vehicle's detour reference speed based on the longitudinal feasible space map; based on the detour reference speed, the feasible space corresponding to the target vehicle's detour behavior stage is sampled and evaluated to determine a driving reference path; based on the longitudinal feasible space map and the driving reference path, the target vehicle's target detour trajectory within a preset time period in the future is determined. Since the present application determines the driving reference path based on the feasible space corresponding to the target vehicle's detour behavior stage based on the detour reference speed, and then determines the target detour trajectory of the target vehicle within a preset time period in the future, so that the target vehicle actively overtakes obstacles according to the target detour trajectory, it can greatly reduce driving risks, improve driving efficiency, and make driving more flexible and comfortable. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0042] Figure 1 A schematic diagram of an application scenario provided in an embodiment of the present application;

[0043] Figure 2 A flowchart of a detour trajectory planning method provided in one embodiment of the present application;

[0044] Figure 3 A flowchart of a detour trajectory planning method provided in another embodiment of the present application;

[0045] Figure 4 A schematic diagram of the structure of a detour trajectory planning device provided in one embodiment of the present application;

[0046] Figure 5 This is a schematic diagram of the structure of an electronic device provided in this application. DETAILED DESCRIPTION

[0047] To make the purpose, technical solutions, and advantages of the embodiments of this application more clear, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0048] In the technical solution of this application, the collection, storage, use, processing, transmission, provision and disclosure of information such as financial data or user data involved comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0049] In complex and ever-changing driving conditions, how to improve driving efficiency, enhance driving flexibility, and improve the driving experience while ensuring driving safety, so that the driving process of autonomous vehicles can be closer to human driving behavior, is one of the topics that need to be studied in autonomous driving trajectory planning.

[0050] At present, during the driving process of an autonomous vehicle, if it senses that an obstacle (such as a slow-moving obstacle such as a pedestrian or non-motor vehicle) has partially entered the lane where the autonomous vehicle is located, the lane of the autonomous vehicle will be blocked and unable to pass. At this time, there are usually two driving decisions for the autonomous vehicle: one is to choose a relatively conservative strategy, that is, to follow behind the obstacle and pass slowly. Although this strategy can greatly reduce driving risks, it has low driving efficiency; the other is to adopt a relatively aggressive overtaking strategy, that is, to find a suitable opportunity and use lane changing to overtake to actively overtake the obstacle, but the above lane changing and overtaking method increases driving risks.

[0051] Based on the above problems, the present application provides a detour trajectory planning method, device, equipment and storage medium. When an obstacle is about to or has partially invaded the lane where the autonomous driving vehicle is located, an active detour and overtaking strategy is adopted on the basis of reserving a sufficient lateral safety threshold. This can greatly reduce driving risks and make driving more flexible and comfortable.

[0052] Below, the application scenarios of the solution provided in this application are first illustrated.

[0053] Figure 1 This is a schematic diagram of an application scenario provided by an embodiment of the present application. Figure 1As shown, in this application scenario, autonomous vehicle 101 is traveling on road 102. When obstacle 103 on road 102 partially intrudes into the lane where autonomous vehicle 101 is located, autonomous vehicle 101 proactively overtakes obstacle 103 according to a planned detour trajectory and continues traveling on road 102. The specific implementation process of how autonomous vehicle 101 obtains the planned detour trajectory can be found in the solutions of the following embodiments.

[0054] It should be noted that Figure 1 This is only a schematic diagram of an application scenario provided by the embodiment of the present application. Figure 1 The equipment included in the Figure 1 The positional relationship between the devices is limited.

[0055] Next, a method for planning a detour trajectory is introduced through a specific embodiment.

[0056] Figure 2 This is a flow chart of a method for planning a detour trajectory provided by an embodiment of the present application. The method of the embodiment of the present application can be applied to an electronic device, which can be a server or a server cluster. Figure 2 As shown, the method of the embodiment of the present application includes:

[0057] S201: Construct a longitudinal feasible space diagram corresponding to the target vehicle circumventing obstacles within a preset time period in the future.

[0058] Among them, the longitudinal feasible space diagram is used to represent the longitudinal feasible range and longitudinal speed constraint information of the target vehicle corresponding to discrete moments within a preset time period in the future.

[0059] In an embodiment of the present application, the future preset time length is, for example, the future 10 seconds. Exemplarily, the horizontal coordinate of the longitudinal feasible space diagram is, for example, a discrete moment within the future preset time length, and the vertical coordinate of the longitudinal feasible space diagram is, for example, a longitudinal feasible range corresponding to each discrete moment, and there is corresponding longitudinal speed constraint information within the longitudinal feasible range. After determining that the target vehicle must bypass an obstacle within the future preset time length, a longitudinal feasible space diagram corresponding to the target vehicle bypassing the obstacle within the future preset time length can be constructed based on the predicted trajectory of the obstacle. As for how to construct the longitudinal feasible space diagram corresponding to the target vehicle bypassing the obstacle within the future preset time length, please refer to the subsequent embodiments and will not be repeated here.

[0060] S202: Obtain a detour reference speed for the target vehicle based on the longitudinal feasible space graph.

[0061] In this step, after obtaining the longitudinal feasible space map, the reference speed of the target vehicle can be obtained based on the longitudinal feasible space map. For details on how to obtain the detour reference speed of the target vehicle based on the longitudinal feasible space map, please refer to the subsequent embodiments and will not be repeated here.

[0062] S203 : Based on the detour reference speed, sample and evaluate the feasible space corresponding to the detour behavior phase of the target vehicle to determine a driving reference path.

[0063] Among them, the detour behavior stage includes the preparation stage before detour, the detour process stage and the stage of returning to the center of the lane after detour is completed. The detour behavior stage is determined based on the longitudinal feasible space window corresponding to the target vehicle's detour around the obstacle within a preset time in the future.

[0064] For example, based on the perception of the target vehicle's driving scene and predicted obstacle information, static or slow-moving obstacles that are about to or have partially intruded into the target vehicle's lane are clustered and treated as an obstacle group. It is reasonably assumed that the obstacle group moves forward at the same speed. Based on the target vehicle's current speed, the fastest longitudinal time at which the obstacle closest to the target vehicle is at the same position is calculated, for example, denoted by t_min. The fastest longitudinal time at which the obstacle farthest from the target vehicle is exceeded is calculated, for example, denoted by t_max. The predicted trajectory of each obstacle within the time range [t_min, t_max] is traversed and queried to obtain the longitudinal and lateral boundary information of each obstacle. The longitudinal minimum boundary value, longitudinal maximum boundary value, lateral minimum boundary value, and lateral maximum boundary value of the obstacle group are continuously updated, ultimately obtaining the longitudinal and lateral ranges of the obstacle group within the time range [t_min, t_max]. Based on the lane width, the lateral range occupied by the obstacle group is eliminated to determine whether there is enough lateral space for the target vehicle to bypass the obstacle, that is, the expanded spatial boundary can ensure that the target vehicle does not completely invade the adjacent lane. The longitudinal feasible space window of the expanded target vehicle is defined as [s_interact_min, s_interact_max] and the lateral feasible space window is defined as [d_interact_min, d_interact_max]. Among them, s_interact_min represents the minimum value of the longitudinal feasible space window of the expanded target vehicle, corresponding to the longitudinal minimum boundary value of the obstacle group; s_interact_max represents the maximum value of the longitudinal feasible space window of the expanded target vehicle, corresponding to the longitudinal maximum boundary value of the obstacle group; d_interact_min represents the minimum value of the lateral feasible space window of the expanded target vehicle; d_interact_max represents the maximum value of the lateral feasible space window of the expanded target vehicle.

[0065] Based on the target vehicle's longitudinal feasible space window, the target vehicle's longitudinal feasible space is divided into three parts according to the detour behavior: preparation before detour, detour process, and return to lane center after detour completion. These correspond to three longitudinally consecutive detour behavior stages: preparation before detour, detour process, and return to lane center after detour completion. The longitudinal feasible range of the target vehicle corresponding to the preparation stage is [0, s_interact_min], the longitudinal feasible range of the target vehicle corresponding to the detour process is [s_interact_min, s_interact_max], and the longitudinal feasible range of the target vehicle corresponding to the return to lane center after detour completion is [s_interact_max, s_path_max], where s_path_max is the preset maximum distance for a single path planning. It can be understood that determining the longitudinal feasible range corresponding to each detour behavior stage, i.e., the longitudinal feasible space corresponding to each detour behavior stage, is determined; and determining the expanded lateral feasible space window of the target vehicle, i.e., the lateral feasible space corresponding to the detour process stage, is determined.

[0066] In this step, after obtaining the detour reference speed, the longitudinal feasible space corresponding to the target vehicle's detour behavior phase, and the lateral feasible space corresponding to the detour process phase, the feasible space corresponding to the target vehicle's detour behavior phase can be sampled and evaluated based on the detour reference speed to determine a driving reference path. For details on how to sample and evaluate the feasible space corresponding to the target vehicle's detour behavior phase based on the detour reference speed to determine the driving reference path, please refer to the subsequent embodiments and will not be further described here.

[0067] S204: Determine a target detour trajectory of the target vehicle within a preset time period in the future based on the longitudinal feasible space map and the driving reference path.

[0068] In this step, after obtaining the longitudinal feasible space map and the driving reference path, a target detour trajectory for the target vehicle within a preset future duration can be determined based on the longitudinal feasible space map and the driving reference path. For details on how to determine the target detour trajectory for the target vehicle within a preset future duration based on the longitudinal feasible space map and the driving reference path, please refer to the subsequent embodiments and will not be further described here.

[0069] After determining the target detour trajectory of the target vehicle within a preset time period in the future, the target vehicle can be controlled to actively overtake obstacles according to the target detour trajectory.

[0070] The detour trajectory planning method provided in the embodiment of the present application constructs a longitudinal feasible space diagram corresponding to the target vehicle's detour around obstacles within a preset time period in the future, and obtains the target vehicle's detour reference speed based on the longitudinal feasible space diagram; based on the detour reference speed, samples and evaluates the feasible space corresponding to the target vehicle's detour behavior phase to determine a driving reference path; and determines the target detour trajectory of the target vehicle within a preset time period in the future based on the longitudinal feasible space diagram and the driving reference path. Since the embodiment of the present application determines the driving reference path based on the feasible space corresponding to the target vehicle's detour behavior phase based on the detour reference speed, and further determines the target detour trajectory of the target vehicle within a preset time period in the future, so that the target vehicle actively overtakes obstacles according to the target detour trajectory, it can greatly reduce driving risks, improve driving efficiency, and make driving more flexible and comfortable.

[0071] Figure 3 This is a flow chart of a detour trajectory planning method provided by another embodiment of the present application. Based on the above embodiment, this embodiment of the present application further explains how to plan a detour trajectory. Figure 3 As shown, the method of the embodiment of the present application may include:

[0072] S301. Cluster obstacles that may occur to the target vehicle within a preset future time period to obtain corresponding clustering results; perform a detour feasibility analysis based on the clustering results; and determine obstacles that the target vehicle can detour within the preset future time period.

[0073] In this embodiment of the present application, for example, referring to the example in step S203, obstacles to the target vehicle within a preset future time period may be, for example, static or slow-moving obstacles that are about to or have already partially intruded into the target vehicle's lane. In this step, based on the perception of the target vehicle's driving scene and predicted obstacle information, these static or slow-moving obstacles that are about to or have already partially intruded into the target vehicle's lane are clustered and treated as an obstacle group. A clustering result corresponding to this obstacle group is then obtained, namely, the longitudinal feasible range and lateral feasible range of the obstacle group within the time range [t_min, t_max]. Based on the clustering results, a detour feasibility analysis is performed. Specifically, first, the longitudinal feasible range of the obstacle group within the time range [t_min, t_max] is defined as the interactive longitudinal area. For this interactive longitudinal area, the lane line attributes of the lane where the target vehicle is located need to be processed. For example, when the lane line attribute is a double yellow line, the interactive longitudinal area is identified as a horizontal non-expandable area, which means that the target vehicle cannot detour around the obstacle within a preset time in the future. Secondly, based on the lane width, the lateral range occupied by the obstacle is eliminated to determine whether there is enough lateral space to detour the obstacle group. The expanded spatial boundary should ensure that the target vehicle does not completely invade the adjacent lane. If there is a reasonable lateral space, the longitudinal feasible space window of the expanded target vehicle is [s_interact_min, s_interact_max] and the lateral feasible space window is [d_interact_min, d_interact_max]. Therefore, the longitudinal feasible space corresponding to each detour behavior stage and the lateral feasible space corresponding to the detour process stage are determined, that is, it is determined that the target vehicle can detour obstacles within the preset time in the future.

[0074] In the embodiment of this application, Figure 2 Step S201 may further include the following two steps S302 and S303:

[0075] S302: Determine a target feasible minimum longitudinal distance, a target feasible maximum longitudinal distance, a minimum target longitudinal speed, and a maximum target longitudinal speed corresponding to the target vehicle circumventing an obstacle at discrete moments within a preset future time period.

[0076] In this step, after determining that the target vehicle can bypass the obstacle within a preset time period in the future, the target longitudinal minimum feasible distance, target longitudinal maximum feasible distance, minimum target longitudinal speed, and maximum target longitudinal speed corresponding to the target vehicle bypassing the obstacle at discrete moments in the preset time period in the future can be determined.

[0077] Further, optionally, determining a target feasible minimum longitudinal distance, a target feasible maximum longitudinal distance, a minimum value of a target longitudinal speed, and a maximum value of a target longitudinal speed corresponding to the target vehicle circumventing an obstacle at discrete moments in the future preset time period may include: determining an initial feasible maximum longitudinal distance, an initial feasible minimum longitudinal distance, a minimum value of an initial longitudinal speed, and a maximum value of an initial longitudinal speed of the target vehicle; and performing the following operations for obstacles to be circumvented at discrete moments in the future preset time period until all obstacles are traversed: determining a target feasible minimum longitudinal distance of the target vehicle based on a longitudinal space occupied by the target obstacle and the initial feasible minimum longitudinal distance, and determining a minimum value of the target longitudinal speed based on the speed of the target obstacle, determining the target feasible minimum longitudinal distance as a new initial feasible minimum longitudinal distance, and determining the minimum value of the target longitudinal speed as a new minimum value of the initial longitudinal speed.

[0078] For example, the maximum feasible distance of the target vehicle in an interference-free state is obtained (expressed as max_drivable_length) for a preset future time period, such as the next 10 seconds. This is the farthest feasible distance of the target vehicle in the next 10 seconds under physical constraints. Considering the target vehicle's driving scenario speed limit, the max_drivable_length is initialized according to the target vehicle's driving scenario speed limit, thereby obtaining the target vehicle's initial longitudinal feasible maximum distance. The target vehicle's initial longitudinal feasible minimum distance is, for example, 0, the target vehicle's initial minimum longitudinal velocity is, for example, 0, and the target vehicle's initial maximum longitudinal velocity is, for example, the target vehicle's driving scenario speed limit.

[0079] After obtaining the target vehicle's initial maximum feasible longitudinal distance, initial minimum feasible longitudinal distance, initial minimum longitudinal velocity, and initial maximum longitudinal velocity, the next 10 seconds are discretized at equal time intervals to obtain the corresponding discrete moments. For each discrete moment (denoted as t_i), the predicted trajectory of the obstacle to be circumvented (denoted as obs_i) is queried. The longitudinal spatial range occupied by the obstacle to be circumvented at discrete moment t_i is calculated as: [obs_i_s_min, obs_i_s_max], where obs_i_s_min represents the minimum longitudinal spatial range occupied by the obstacle to be circumvented, and obs_i_s_max represents the maximum longitudinal spatial range occupied by the obstacle to be circumvented. This spatial range represents the target vehicle's impassable zone. Considering that the target vehicle needs to circumvent the obstacle, i.e., an overtaking process, the target minimum feasible longitudinal distance of the target vehicle at discrete moment t_i is obs_i_s_max. And because it is a non-reversing trajectory planning, it is necessary to ensure that the target longitudinal minimum feasible distance at discrete time t_i is not less than the target longitudinal minimum feasible distance at the previous discrete time, and the entire process is constrained by the maximum feasible distance. If the obstacle is close to the target vehicle, the minimum target longitudinal speed of the target vehicle needs to be updated to the speed of the obstacle. For each discrete moment in the next 10 seconds to be bypassed, starting from discrete time t = 0s, the calculation is iterated step by step according to a fixed time resolution to t = 10s to obtain the longitudinal feasible longitudinal range (corresponding to the target longitudinal minimum feasible distance and the target longitudinal maximum feasible distance) and longitudinal speed constraint information (corresponding to the minimum target longitudinal speed and the maximum target longitudinal speed) at each discrete time t_i.

[0080] S303: Construct a longitudinal feasible space diagram according to the target longitudinal feasible minimum distance, the target longitudinal feasible maximum distance, the minimum value of the target longitudinal speed, and the maximum value of the target longitudinal speed.

[0081] In this step, illustratively, the future preset time length is, for example, the next 10 seconds. After obtaining the target longitudinal feasible minimum distance, the target longitudinal feasible maximum distance, the minimum value of the target longitudinal speed, and the maximum value of the target longitudinal speed, a longitudinal feasible space diagram of the target vehicle corresponding to the next 10 seconds can be constructed based on the target longitudinal feasible minimum distance, the target longitudinal feasible maximum distance, the minimum value of the target longitudinal speed, and the maximum value of the target longitudinal speed. The diagram is used to describe the longitudinal feasible minimum distance and the longitudinal feasible maximum distance corresponding to the target vehicle at each discrete moment in the next 10 seconds, that is, the longitudinal feasible range of the target vehicle corresponding to each discrete moment in the next 10 seconds, and within the longitudinal feasible range there is a corresponding minimum value of the target longitudinal speed and a maximum value of the target longitudinal speed, that is, the longitudinal speed constraint information.

[0082] In the embodiment of this application, Figure 2 Step S202 may further include the following four steps S304 to S307:

[0083] S304. Based on the longitudinal feasible space graph, according to the maximum acceleration capability, maximum deceleration capability of the target vehicle, the curvature corresponding to the reference line of the path within the longitudinal feasible range, and the correspondence between the preset curvature and the speed constraint, obtain updated longitudinal feasible range and longitudinal speed constraint information.

[0084] In this step, the correspondence between the preset curvature and speed constraints is, for example, a curvature-speed constraint table pre-calibrated offline, and the reference line of the path within the longitudinal feasible range is, for example, the center line of the path within the longitudinal feasible range. For example, the following three constraints are first determined: (1) Considering the driving experience and the continuity of the controller control signal of the target vehicle (such as the control signal corresponding to the opening of the accelerator pedal), it is necessary to ensure the continuity of the initial state of the target vehicle (i.e., speed and acceleration), that is, the planned initial speed and initial acceleration at the starting point are initialized according to the current execution state of the target vehicle; (2) Considering the execution capability of the target vehicle, such as the maximum acceleration capability and maximum deceleration capability of the target vehicle, determine the constraint information of the executable speed at the next discrete moment and the constraint information of the executable distance at the next discrete moment; (3) Considering the lane shape, such as excessive speed during turning can easily cause the target vehicle to become unstable or reduce the riding experience, therefore, it is necessary to add constraints to the speed of the target vehicle according to the curvature of the lane line; specifically, for example, for each discrete moment t_i, any sampling point that meets the boundary value constraint of the longitudinal feasible range is represented as (t_i, s_i), where s_i represents the longitudinal feasible distance of any sampling point, and the corresponding curvature information on the reference line is queried, and the maximum speed constraint at the position is obtained by querying the correspondence between the preset curvature and the speed constraint.

[0085] Based on the longitudinal feasible range and longitudinal speed constraint information corresponding to discrete moments of the target vehicle within a preset future time range in the longitudinal feasible space diagram, the updated longitudinal feasible range and longitudinal speed constraint information can be obtained according to the above three constraints.

[0086] S305 : Obtain an updated longitudinal feasible space diagram according to the updated longitudinal feasible range and longitudinal speed constraint information.

[0087] In this step, after obtaining the updated longitudinal feasible range and longitudinal speed constraint information corresponding to each discrete moment within the future preset time period, an updated longitudinal feasible space diagram can be obtained based on the updated longitudinal feasible range and longitudinal speed constraint information.

[0088] S306 : Based on the updated longitudinal feasible space graph and the second preset cost function, obtain a first target position of the target vehicle corresponding to a discrete moment.

[0089] The second preset cost function is used to express the quantitative relationship between the cost value and the velocity, acceleration and jerk of the corresponding discrete moment sampling point.

[0090] For example, a too low driving speed will lead to a low driving efficiency problem, while a too fast speed will easily affect the driving experience, and in severe cases, it may even cause vehicle instability. Therefore, when defining the second preset cost function, it is necessary to balance the driving efficiency and comfort of the target vehicle, that is, for the sake of driving efficiency, the driving distance tends to be as far as possible; for the sake of comfort, the driving process is expected to be smoother, and the acceleration and jerk (i.e., the increment of acceleration) tend to be as small as possible. Therefore, the second preset cost function is defined as a quantitative relationship between the cost value and the speed, acceleration, and jerk of the sampling point at the corresponding discrete moment. In this step, after obtaining the updated longitudinal feasible space graph, the first target position of the target vehicle corresponding to each discrete moment can be obtained based on the updated longitudinal feasible space graph and the second preset cost function. This process can be understood as a dynamic programming solution process in a finite space. Finally, a discrete sequence of boundary values that meet the longitudinal feasible range can be searched and increased in time, represented as the st discrete sequence, where s represents the first target position of the target vehicle corresponding to the discrete moment t.

[0091] S307: Perform fitting processing on the first target position to obtain a reference speed of the target vehicle.

[0092] In this step, after obtaining the first target position, a fitting process can be performed on the first target position to obtain a reference speed of the target vehicle. For example, the st discrete sequence points obtained in step S306 are fitted to obtain the reference speed of the target vehicle. The reference speed is specifically described in the form of a position-time curve.

[0093] In the embodiment of this application, Figure 2 The step S203 may further include the following four steps S308 to S311:

[0094] S308 : Sampling the feasible space corresponding to the detour behavior phase of the target vehicle to obtain corresponding node information.

[0095] For example, a node is the smallest unit in a path optimization solution. The node stores the corresponding sampling point and its connection information with the corresponding forward or backward node. In this step, the target vehicle's detour behavior phases include the pre-detour preparation phase, the detour process phase, and the post-detour return to the lane center phase. Each detour behavior phase has a corresponding feasible space. Therefore, the feasible space corresponding to each detour behavior phase of the target vehicle can be sampled to obtain the corresponding node information.

[0096] Furthermore, optionally, sampling the feasible space corresponding to the bypass behavior stage of the target vehicle to obtain corresponding node information can include: sampling the longitudinal feasible space of each bypass behavior stage at equal intervals to obtain the corresponding discrete longitudinal feasible distance; sampling the lateral feasible space of the target vehicle at equal intervals based on the discrete longitudinal feasible distance, lane width, and position information of the obstacle to be bypassed to obtain corresponding sampling points; and obtaining corresponding node information based on the sampling points.

[0097] Exemplarily, the longitudinal feasible space of each detour behavior stage is sampled at equal intervals to obtain the corresponding discrete longitudinal feasible distance. For each discrete longitudinal feasible distance, the lane width information can be queried to obtain the lateral feasible space of the target vehicle corresponding to the discrete longitudinal feasible distances contained in the preparation stage before the detour and the stage of returning to the center of the lane after the detour is completed; for the discrete longitudinal feasible distances contained in the detour process stage, the corresponding lateral feasible space of the target vehicle can be obtained according to the lateral feasible space window [d_interact_min, d_interact_max] determined in step S301. The lateral feasible space is obtained based on the lane width and the position information of the obstacle to be detoured. It can be understood that each discrete longitudinal feasible distance and the corresponding lateral feasible space of the target vehicle constitute a corresponding sampling layer. Generally, each detour behavior stage corresponds to multiple sampling layers.

[0098] The target vehicle's lateral feasible space is sampled at equal intervals to obtain corresponding sampling points. Each sampling point stores lateral and longitudinal information about the relative position of the lane (represented by sd_state), as well as attribute information indicating whether it can be extended directly forward along the reference line (represented by extendable). Sampling points in the detour process stage have the attribute of extending directly forward along the reference line, and the end point of the forward extension is the maximum distance corresponding to that detour process stage. Therefore, each detour behavior stage of the target vehicle corresponds to a longitudinal feasible range. Each detour behavior stage contains a discrete continuous sampling layer, which corresponds to the longitudinal position and lateral feasible range. The sampling layer contains discrete sampling points, each of which stores sd_state information, index information corresponding to each detour behavior stage, attribute information indicating whether the sampling point contained in each detour behavior stage can be extended directly forward along the reference line, and information on the maximum distance corresponding to each detour behavior stage. Based on the sampling points, the corresponding node information is obtained. It can be understood that there is a one-to-one correspondence between sampling points and nodes.

[0099] S309 : Determine a trajectory corresponding to the target edge according to the detour reference speed and the edge information, wherein the edge information is obtained by connecting nodes included in two adjacent detour behavior stages.

[0100] Exemplarily, an edge is obtained by connecting the nodes included in two adjacent detour behavior stages, thus including forward and backward nodes. The edge construction process needs to follow the following rules: the longitudinal feasible distance corresponding to the longitudinal position of the forward node must be less than the longitudinal feasible distance corresponding to the longitudinal position of the backward node; any two nodes in the same detour behavior stage do not have the attributes of a forward or backward node and cannot be connected, while a node in a detour behavior stage, when used as a forward node, can be connected to any node in the next detour behavior stage; wherein, if the sampling point corresponding to a node has the attribute of being able to extend directly forward along the reference line direction, then when used as a forward node, in addition to being able to connect to the sampling point in the next detour behavior stage, the node can also be extended along the reference line direction to the farthest distance corresponding to the current detour behavior stage. This extension point is then used as a new sampling point to connect to any node in the next detour behavior stage, forming a new edge. In this step, after obtaining the detour reference speed and edge information, the trajectory corresponding to the target edge can be determined based on the detour reference speed and edge information.

[0101] Further, optionally, determining the trajectory corresponding to the target edge based on the detour reference speed and the edge information may include: discretizing and interpolating the longitudinal feasible distance range corresponding to the edge based on the detour reference speed to determine the trajectory corresponding to the target edge.

[0102] For example, for each connected edge, based on the longitudinal feasible distance range corresponding to the edge, a position-time curve segment corresponding to the longitudinal feasible distance range is intercepted from the reference velocity. Through discretization and interpolation, the trajectory corresponding to the target edge is determined. The trajectory contains trajectory points corresponding to information such as position, velocity, acceleration, and heading angle.

[0103] S310: Obtain a cost value of the trajectory corresponding to the target edge according to a first preset cost function.

[0104] The first preset cost function is used to express a quantitative relationship between the cost value and the trajectory difference, speed, collision risk, and maximum yaw angular velocity of the corresponding trajectory.

[0105] In this step, the cost can also be called expenditure. In trajectory planning, when defining the first preset cost function, the following four factors need to be considered: (1) consistency, which represents the degree of similarity with the execution trajectory of the previous step. Due to the continuity of the driving control process, the difference in trajectory directly affects the change amplitude of the steering wheel control signal. In order to ensure the consistency and stability of the trajectory planning results, the execution paths before and after should not deviate too much; (2) collision risk, based on the predicted trajectory of the obstacle, the potential collision risk of the candidate trajectory and the obstacle in the driving scene of the target vehicle is calculated. Different safety thresholds are reserved for static obstacles and dynamic obstacles respectively, because the collision of static obstacles at the path level is inevitable, while the collision risk of dynamic obstacles can be eliminated by trying to adjust the speed; (3) stability, considering the execution capability and execution effect of the controller, it is necessary to have certain constraints on the yaw rate of the trajectory to avoid excessive yaw rate causing vehicle instability; (4) driving tendency, which represents the tendency of the driving process. The slower the trajectory speed, the higher the stability, but it will cause low driving efficiency. Therefore, driving tendency is a trade-off between the stability of the trajectory and the execution efficiency. Under the premise of ensuring stability, the vehicle speed is increased, thereby improving driving efficiency. Based on the above four factors, the first preset cost function is defined as a quantitative relationship between the cost value and the trajectory difference, speed, collision risk and maximum yaw rate of the corresponding trajectory.

[0106] In this step, after the first preset cost function is determined, the cost value of the trajectory corresponding to the target edge can be obtained according to the first preset cost function.

[0107] S311. Determine a driving reference path according to the cost value.

[0108] In this step, after obtaining the cost value of the trajectory corresponding to the target edge, a driving reference path can be determined based on the cost value of the trajectory corresponding to the target edge. The driving reference path can also be understood as an optimal path.

[0109] Furthermore, optionally, a driving reference path is determined based on the cost value, including: taking the initial position of the target vehicle as the root node, connecting each edge included in the detour behavior stage from the root node in sequence to obtain a node connection sequence; determining the node connection sequence with the minimum cost value addition; and determining the driving reference path based on the edge corresponding to the node connection sequence with the minimum cost value addition.

[0110] Exemplarily, the search uses the initial position of the target vehicle as the root node, connects to each edge from the root node, uses the backward node as the forward node, and then connects to each edge of the next detour behavior stage, until the node of the edge does not have a corresponding backward node. For each node of each connection method, traverse each forward node connected to the edge, and find the node connection sequence with the smallest cost sum, which is the optimal solution. Connect the edges corresponding to the node connection sequence with the smallest cost sum in sequence to obtain a driving reference path (i.e., the optimal path), which is specifically described in the form of a driving reference path curve. It can be understood that the node stores the forward node with the minimum cost to reach the node and the corresponding minimum cost value information.

[0111] S312: Update the longitudinal speed constraint information of the longitudinal feasible space graph according to the driving reference path, the location information and the type attribute of the obstacle.

[0112] Exemplarily, the longitudinal position information of the obtained driving reference path is discretely sampled, and the lateral position information is obtained based on the interpolation of the driving reference path curve, which represents the lateral offset of the center of the target vehicle relative to the reference line; at the same time, the position information of the obstacle is queried, and the minimum lateral space between the target vehicle and the obstacle is calculated. The speed constraint is set based on the type attribute of the obstacle (such as pedestrian, motor vehicle or non-motor vehicle, etc.). If the target vehicle is too close to the obstacle, it is necessary to actively slow down and pass carefully; if it is outside the safe range, it can pass at a normal speed, and the longitudinal speed constraint information of the longitudinal feasible space diagram of the target vehicle is re-updated to narrow the solution space, refine the speed solution space, and thereby improve the solution accuracy and efficiency of speed optimization.

[0113] In the embodiment of this application, Figure 2 The step S204 may further include the following step S313:

[0114] S313. Based on the updated longitudinal feasible space graph, speed sampling is performed to obtain sampled alternative speeds; based on the sampled alternative speeds and the driving reference path, an alternative detour trajectory is generated; based on the alternative detour trajectory and a third preset cost function, a target detour trajectory is determined.

[0115] The third preset cost function is used to express a quantitative relationship between the cost value and the collision risk, maximum acceleration, maximum jerk, and maximum yaw rate of the corresponding alternative trajectory.

[0116] For example, when defining the third preset cost function, it is necessary to comprehensively consider the potential collision risk of the alternative trajectory, that is, the probability that the target vehicle and the obstacle appear in the same position at the same time; the comfort of the trajectory, that is, the acceleration of the trajectory and its rate of change; and the stability of the trajectory, that is, the maximum yaw rate of the trajectory. Therefore, the third preset cost function is defined as a quantitative relationship between the cost value and the collision risk, maximum acceleration, maximum jerk, and maximum yaw rate of the corresponding alternative trajectory. In this step, speed sampling is performed within the updated longitudinal feasible space graph to obtain the sampled alternative speed. Based on the sampled alternative speed, an alternative detour trajectory is generated in combination with the driving reference path. For each alternative detour trajectory, the target detour trajectory is determined from the alternative detour trajectories according to the third preset cost function, and the target detour trajectory is published to the control module to control the target vehicle to travel according to the planned detour trajectory.

[0117] The detour trajectory planning method provided by the embodiment of the present application obtains corresponding clustering processing results by clustering obstacles encountered by the target vehicle within a preset time period in the future; performs detour feasibility analysis based on the clustering processing results; determines obstacles that the target vehicle can detour within a preset time period in the future; determines the target longitudinal feasible minimum distance, target longitudinal feasible maximum distance, target longitudinal speed minimum value and target longitudinal speed maximum value corresponding to the target vehicle detour obstacles at discrete moments within the preset time period in the future; constructs a longitudinal feasible space diagram based on the target longitudinal feasible minimum distance, target longitudinal feasible maximum distance, target longitudinal speed minimum value and target longitudinal speed maximum value; based on the longitudinal feasible space diagram, obtains updated longitudinal feasible range and longitudinal speed constraint information according to the maximum acceleration capability, maximum deceleration capability of the target vehicle, the curvature corresponding to the reference line of the path within the longitudinal feasible range and the correspondence between the preset curvature and the speed constraint; obtains updated longitudinal feasible range and longitudinal speed constraint information based on the updated longitudinal feasible range the longitudinal feasible space diagram is obtained based on the travel range and longitudinal speed constraint information; based on the updated longitudinal feasible space diagram and the second preset cost function, the first target position of the target vehicle corresponding to the discrete moment is obtained; the first target position is fitted to obtain the reference speed of the target vehicle; the feasible space corresponding to the detour behavior stage of the target vehicle is sampled to obtain the corresponding node information; according to the detour reference speed and edge information, the trajectory corresponding to the target edge is determined; according to the first preset cost function, the cost value of the trajectory corresponding to the target edge is obtained; according to the cost value, the driving reference path is determined; according to the driving reference path, the position information and type attributes of the obstacle, the longitudinal speed constraint information of the longitudinal feasible space diagram is updated; based on the updated longitudinal feasible space diagram, speed sampling is performed to obtain the sampled alternative speeds; according to the sampled alternative speeds and the driving reference path, the alternative detour trajectory is generated; according to the alternative detour trajectory and the third preset cost function, the target detour trajectory is determined. Since the embodiment of the present application is based on the detour reference speed, the driving reference path is determined according to the feasible space corresponding to the detour behavior stage of the target vehicle, and the longitudinal speed constraint information of the longitudinal feasible space diagram is updated according to the driving reference path and the position information and type attributes of the obstacle. Then, based on the updated longitudinal feasible space diagram and the driving reference path, the target detour trajectory of the target vehicle within a preset time period in the future is determined, so that the target vehicle actively overtakes the obstacle according to the target detour trajectory. Therefore, it can greatly reduce driving risks, improve driving efficiency, make driving more flexible and comfortable, and improve the efficiency of determining the target detour trajectory.

[0118] The following are device embodiments of the present application, which can be used to implement the method embodiments of the present application. For details not disclosed in the device embodiments of the present application, please refer to the method embodiments of the present application.

[0119] Figure 4 This is a schematic diagram of the structure of the detour trajectory planning device provided in one embodiment of the present application, as shown in FIG. Figure 4 As shown, the detour trajectory planning device 400 of the embodiment of the present application includes: a construction module 401, an acquisition module 402, a processing module 403 and a first determination module 404. Among them:

[0120] Construction module 401 is used to construct a longitudinal feasible space diagram corresponding to the target vehicle's detour around obstacles within a preset future time period. The longitudinal feasible space diagram is used to represent the longitudinal feasible range and longitudinal speed constraint information corresponding to discrete moments in the future preset time period.

[0121] The acquisition module 402 is configured to obtain a detour reference speed of the target vehicle based on the longitudinal feasible space map.

[0122] Processing module 403 is used to sample and evaluate the feasible space corresponding to the bypass behavior stage of the target vehicle based on the bypass reference speed to determine the driving reference path. The bypass behavior stage includes the preparation stage before the bypass, the bypass process stage, and the stage of returning to the center of the lane after the bypass is completed. The bypass behavior stage is determined based on the longitudinal feasible space window corresponding to the target vehicle's bypass of the obstacle within a preset time period in the future.

[0123] The first determining module 404 is configured to determine a target detour trajectory of the target vehicle within a preset time period in the future according to the longitudinal feasible space graph and the driving reference path.

[0124] In some embodiments, the processing module 403 can be specifically used to: sample the feasible space corresponding to the detour behavior stage of the target vehicle to obtain corresponding node information; determine the trajectory corresponding to the target edge based on the detour reference speed and edge information, and the edge information is the edge information obtained by connecting the nodes included in two adjacent detour behavior stages; obtain the cost value of the trajectory corresponding to the target edge according to a first preset cost function, and the first preset cost function is used to express the quantitative relationship between the cost value and the trajectory difference, speed, collision risk and maximum yaw angular velocity of the corresponding trajectory; determine the driving reference path based on the cost value.

[0125] Optionally, when the processing module 403 is used to sample the feasible space corresponding to the bypass behavior stage of the target vehicle and obtain the corresponding node information, it can be specifically used to: perform equidistant sampling on the longitudinal feasible space of each bypass behavior stage to obtain the corresponding discrete longitudinal feasible distance; perform equidistant sampling on the lateral feasible space of the target vehicle based on the discrete longitudinal feasible distance, lane width and position information of the obstacle to be bypassed to obtain the corresponding sampling points; and obtain the corresponding node information based on the sampling points.

[0126] Optionally, when the processing module 403 is used to determine the trajectory corresponding to the target edge based on the detour reference speed and edge information, it can be specifically used to: discretize and interpolate the longitudinal feasible distance range corresponding to the edge based on the detour reference speed to determine the trajectory corresponding to the target edge.

[0127] Optionally, when the processing module 403 is used to determine the driving reference path based on the cost value, it can be specifically used to: take the initial position of the target vehicle as the root node, connect each edge included in the detour behavior stage from the root node in sequence to obtain a node connection sequence; determine the node connection sequence with the minimum cost value addition; and determine the driving reference path based on the edge corresponding to the node connection sequence with the minimum cost value addition.

[0128] In some embodiments, the construction module 401 can be specifically used to: determine the target longitudinal feasible minimum distance, target longitudinal feasible maximum distance, minimum target longitudinal speed, and maximum target longitudinal speed corresponding to the target vehicle's circumvention of obstacles at discrete moments within a preset time period in the future; and construct a longitudinal feasible space diagram based on the target longitudinal feasible minimum distance, target longitudinal feasible maximum distance, minimum target longitudinal speed, and maximum target longitudinal speed.

[0129] Optionally, when the construction module 401 is used to determine the target feasible minimum longitudinal distance, target feasible maximum longitudinal distance, minimum target longitudinal speed, and maximum target longitudinal speed corresponding to the target vehicle circumventing obstacles at discrete moments in the future preset time period, it can be specifically used to: determine the target vehicle's initial feasible maximum longitudinal distance, initial feasible minimum longitudinal distance, minimum initial longitudinal speed, and maximum initial longitudinal speed; and for obstacles to be circumvented at discrete moments in the future preset time period, perform the following operations until all obstacles are traversed: determine the target feasible minimum longitudinal distance of the target vehicle based on the longitudinal space occupied by the target obstacle and the initial feasible minimum longitudinal distance, determine the minimum value of the target longitudinal speed based on the speed of the target obstacle, determine the target feasible minimum longitudinal distance as the new initial feasible minimum longitudinal distance, and determine the minimum value of the target longitudinal speed as the new minimum value of the initial longitudinal speed.

[0130] In some embodiments, the acquisition module 402 can be specifically used to: based on the longitudinal feasible space diagram, according to the maximum acceleration capability, maximum deceleration capability, the curvature corresponding to the reference line of the path within the longitudinal feasible range, and the correspondence between the preset curvature and the speed constraint, obtain the updated longitudinal feasible range and longitudinal speed constraint information; based on the updated longitudinal feasible range and longitudinal speed constraint information, obtain the updated longitudinal feasible space diagram; based on the updated longitudinal feasible space diagram and the second preset cost function, obtain the first target position of the target vehicle corresponding to the discrete moment, the second preset cost function is used to express the quantitative relationship between the cost value and the speed, acceleration and jerk of the sampling point corresponding to the discrete moment; fit the first target position to obtain the reference speed of the target vehicle.

[0131] In some embodiments, the first determination module 404 can be specifically used to: perform speed sampling based on the longitudinal feasible space diagram to obtain sampled alternative speeds; generate alternative detour trajectories based on the sampled alternative speeds and the driving reference path; determine the target detour trajectory based on the alternative detour trajectory and a third preset cost function, where the third preset cost function is used to represent the quantitative relationship between the cost value and the collision risk, maximum acceleration, maximum jerk, and maximum yaw angular velocity of the corresponding alternative trajectory.

[0132] Optionally, the detour trajectory planning device 400 also includes an updating module 405, which is used to update the longitudinal speed constraint information of the longitudinal feasible space diagram according to the driving reference path, the position information and type attributes of the obstacle before the first determination module 404 performs speed sampling based on the longitudinal feasible space diagram.

[0133] Optionally, the detour trajectory planning device 400 also includes a second determination module 406, which is used to cluster the obstacles of the target vehicle within a future preset time period before the construction module 401 constructs the longitudinal feasible space diagram corresponding to the target vehicle's detour around obstacles within a future preset time period to obtain corresponding clustering processing results; perform detour feasibility analysis based on the clustering processing results; and determine that the target vehicle can detour around obstacles within a future preset time period.

[0134] The device of this embodiment can be used to execute the technical solution of any of the above-mentioned method embodiments. Its implementation principles and technical effects are similar and will not be repeated here.

[0135] Figure 5 This is a schematic diagram of the structure of an electronic device provided in this application. Figure 5 As shown, the electronic device 500 may include: at least one processor 501 and a memory 502.

[0136] The memory 502 is used to store programs. Specifically, the programs may include program codes, and the program codes include computer operation instructions.

[0137] The memory 502 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0138] The processor 501 is used to execute the computer-executable instructions stored in the memory 502 to implement the detour trajectory planning method described in the aforementioned method embodiment. Among them, the processor 501 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application. Specifically, when implementing the detour trajectory planning method described in the aforementioned method embodiment, the electronic device may be, for example, an electronic device with processing functions such as a terminal and a server. When implementing the detour trajectory planning method described in the aforementioned method embodiment, the electronic device may be, for example, an electronic control unit on a vehicle.

[0139] Optionally, the electronic device 500 may further include a communication interface 503. In a specific implementation, if the communication interface 503, the memory 502, and the processor 501 are implemented independently, the communication interface 503, the memory 502, and the processor 501 may be interconnected via a bus and communicate with each other. The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be divided into address buses, data buses, control buses, etc., but this does not mean that there is only one bus or only one type of bus.

[0140] Optionally, in a specific implementation, if the communication interface 503, the memory 502 and the processor 501 are integrated on a chip, the communication interface 503, the memory 502 and the processor 501 can complete communication through an internal interface.

[0141] The present application also provides a computer-readable storage medium, which may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, and other media that can store program codes. Specifically, the computer-readable storage medium stores program instructions, and the program instructions are used for the detour trajectory planning method in the above embodiment.

[0142] The present application also provides a computer program product including execution instructions stored in a readable storage medium. At least one processor of an electronic device can read the execution instructions from the readable storage medium, and the at least one processor executes the execution instructions to cause the electronic device to implement the detour trajectory planning methods provided in the various embodiments described above.

[0143] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A detour trajectory planning method, characterized in that: include: Constructing a longitudinal feasible space diagram corresponding to the target vehicle's detour around obstacles within a preset future time period, wherein the longitudinal feasible space diagram is used to represent the longitudinal feasible range and longitudinal speed constraint information of the target vehicle at discrete moments within the preset future time period; Based on the longitudinal feasible space map, obtaining a detour reference speed of the target vehicle; Based on the detour reference speed, sampling and evaluating the feasible space corresponding to the detour behavior phase of the target vehicle to determine a driving reference path, wherein the detour behavior phase includes a pre-detour preparation phase, a detour process phase, and a lane center return phase after detour completion. The detour behavior phase is determined based on the longitudinal feasible space window corresponding to the target vehicle detour the obstacle within the future preset time period; The step of sampling and evaluating the feasible space corresponding to the detour behavior phase of the target vehicle based on the detour reference speed to determine a driving reference path includes: Sampling the feasible space corresponding to the detour behavior phase of the target vehicle to obtain corresponding node information; determining a trajectory corresponding to a target edge based on the detour reference speed and edge information, the edge information being edge information obtained by connecting nodes included in two adjacent detour behavior stages; obtaining a cost value of the trajectory corresponding to the target edge based on a first preset cost function, the first preset cost function being used to represent a quantitative relationship between the cost value and a trajectory difference, speed, collision risk, and maximum yaw rate of the corresponding trajectory; and determining a driving reference path based on the cost value; A target detour trajectory of the target vehicle within the future preset time period is determined according to the longitudinal feasible space graph and the driving reference path.

2. The method for planning a detour trajectory according to claim 1, wherein: The sampling of the feasible space corresponding to the detour behavior phase of the target vehicle to obtain corresponding node information includes: Performing equal-distance sampling on the longitudinal feasible space of each detour behavior stage to obtain a corresponding discrete longitudinal feasible distance; Based on the discrete longitudinal feasible distance, lane width, and location information of the obstacle to be circumvented, sampling the lateral feasible space of the target vehicle at equal intervals to obtain corresponding sampling points; According to the sampling point, corresponding node information is obtained.

3. The method for planning a detour trajectory according to claim 1, wherein: The determining, based on the detour reference speed and the edge information, a trajectory corresponding to the target edge includes: According to the detour reference speed, the longitudinal feasible distance range corresponding to the edge is discretized and interpolated to determine the trajectory corresponding to the target edge.

4. The method for planning a detour trajectory according to claim 1, wherein: The determining of the driving reference path according to the cost value includes: Taking the initial position of the target vehicle as a root node, sequentially connecting each edge included in the detour behavior stage from the root node to obtain a node connection sequence; Determine the node connection sequence with the minimum sum of cost; A driving reference path is determined based on the edge corresponding to the node connection sequence with the smallest sum of the cost values.

5. The method for planning a detour trajectory according to any one of claims 1 to 4, characterized in that: The step of constructing a longitudinal feasible space diagram corresponding to the target vehicle circumventing obstacles within a preset time period in the future includes: Determining a target feasible minimum longitudinal distance, a target feasible maximum longitudinal distance, a minimum target longitudinal speed, and a maximum target longitudinal speed corresponding to the target vehicle circumventing the obstacle at discrete moments within the future preset time period; The longitudinal feasible space diagram is constructed according to the target longitudinal feasible minimum distance, the target longitudinal feasible maximum distance, the minimum value of the target longitudinal speed, and the maximum value of the target longitudinal speed.

6. The method for planning a detour trajectory according to claim 5, characterized in that: The determining of a target feasible minimum longitudinal distance, a target feasible maximum longitudinal distance, a minimum target longitudinal speed, and a maximum target longitudinal speed corresponding to the target vehicle circumventing the obstacle at discrete moments within the future preset time period includes: Determining an initial feasible maximum longitudinal distance, an initial feasible minimum longitudinal distance, a minimum initial longitudinal velocity, and a maximum initial longitudinal velocity of the target vehicle; For obstacles to be circumvented at discrete moments within the future preset time period, perform the following operations until all obstacles are traversed: The target longitudinal feasible minimum distance of the target vehicle is determined based on the longitudinal space occupied by the target obstacle and the initial longitudinal feasible minimum distance, and the minimum value of the target longitudinal speed is determined based on the speed of the target obstacle. The target longitudinal feasible minimum distance is determined as the new initial longitudinal feasible minimum distance, and the minimum value of the target longitudinal speed is determined as the new minimum value of the initial longitudinal speed.

7. The method for planning a detour trajectory according to any one of claims 1 to 4, characterized in that: The obtaining, based on the longitudinal feasible space diagram, a detour reference speed of the target vehicle includes: Based on the longitudinal feasible space graph, obtaining updated longitudinal feasible range and longitudinal speed constraint information according to the maximum acceleration capability and maximum deceleration capability of the target vehicle, the curvature corresponding to the reference line of the path within the longitudinal feasible range, and the correspondence between the preset curvature and the speed constraint; Obtaining an updated longitudinal feasible space diagram according to the updated longitudinal feasible range and longitudinal speed constraint information; Obtaining a first target position of the target vehicle corresponding to the discrete moment based on the updated longitudinal feasible space graph and a second preset cost function, wherein the second preset cost function is used to represent a quantitative relationship between a cost value and a velocity, acceleration, and jerk of a sampling point corresponding to the discrete moment; A fitting process is performed on the first target position to obtain a reference speed of the target vehicle.

8. The method for planning a detour trajectory according to any one of claims 1 to 4, characterized in that: The determining, based on the longitudinal feasible space graph and the driving reference path, a target detour trajectory of the target vehicle within the future preset time period includes: Based on the longitudinal feasible space map, speed sampling is performed to obtain sampled alternative speeds; generating an alternative detour trajectory according to the sampled alternative speed and the driving reference path; The target bypass trajectory is determined based on the alternative bypass trajectory and a third preset cost function, where the third preset cost function is used to represent a quantitative relationship between a cost value and the collision risk, maximum acceleration, maximum jerk, and maximum yaw rate of the corresponding alternative trajectory.

9. The method for planning a detour trajectory according to claim 8, wherein: Before performing speed sampling based on the longitudinal feasible space map, the method further includes: The longitudinal speed constraint information of the longitudinal feasible space graph is updated according to the driving reference path, the position information and the type attribute of the obstacle.

10. The method for planning a detour trajectory according to any one of claims 1 to 4, characterized in that: Before constructing the longitudinal feasible space map corresponding to the target vehicle circumventing obstacles within a preset time period in the future, the method further includes: Performing clustering processing on obstacles of the target vehicle within the future preset time period to obtain corresponding clustering processing results; Performing a detour feasibility analysis based on the clustering processing results; Determine whether the target vehicle can bypass the obstacle within the future preset time period.

11. A detour trajectory planning device, characterized in that: include: A construction module is used to construct a longitudinal feasible space diagram corresponding to the target vehicle's detour around obstacles within a preset future time period, wherein the longitudinal feasible space diagram is used to represent the longitudinal feasible range and longitudinal speed constraint information of the target vehicle at discrete moments within the preset future time period; An acquisition module, configured to obtain a detour reference speed of the target vehicle based on the longitudinal feasible space map; a processing module configured to sample and evaluate feasible spaces corresponding to detour behavior phases of the target vehicle based on the detour reference speed to determine a driving reference path, wherein the detour behavior phases include a pre-detour preparation phase, a detour process phase, and a lane center return phase after detour completion, and wherein the detour behavior phases are determined based on a longitudinal feasible space window corresponding to the target vehicle detour around the obstacle within the future preset time period; The processing module is specifically configured to sample the feasible space corresponding to the detour behavior phase of the target vehicle to obtain corresponding node information; determine a trajectory corresponding to the target edge based on the detour reference speed and edge information, wherein the edge information is edge information obtained by connecting two adjacent nodes included in the detour behavior phase; obtain a cost value of the trajectory corresponding to the target edge based on a first preset cost function, wherein the first preset cost function is configured to represent a quantitative relationship between the cost value and a trajectory difference, speed, collision risk, and maximum yaw angular velocity of the corresponding trajectory; and determine a driving reference path based on the cost value; A determination module is used to determine a target detour trajectory of the target vehicle within the future preset time period based on the longitudinal feasible space diagram and the driving reference path.

12. An electronic device, characterized in that: include: a processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the bypass trajectory planning method according to any one of claims 1 to 10.

13. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the method for planning a bypass trajectory according to any one of claims 1 to 10 is implemented.

14. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the bypass trajectory planning method according to any one of claims 1 to 10 is implemented.

Images