Parking trajectory planning method, storage medium and vehicle

Abstract

The present disclosure relates to a parking trajectory planning method, a storage medium and a vehicle. The parking trajectory planning method comprises: determining a plurality of extension positions according to a relative position relationship between a starting position of a vehicle and a parking position, wherein a drivable trajectory exists from the parking position to the extension positions; and determining a planning trajectory from the starting position to the parking position according to the plurality of extension positions. The parking trajectory planning method, the storage medium and the vehicle can realize efficient and high-precision parking trajectory planning.

Description

Technical Field

[0001] This disclosure relates to the field of vehicle technology, specifically to a parking trajectory planning method, a storage medium, and a vehicle. Background Technology

[0002] With the development of vehicle technology, vehicles can park using planned parking trajectories. The quality of the parking trajectory determines whether a vehicle can park successfully.

[0003] The trajectory search algorithm used by the vehicle can determine the quality of the parking trajectory. Summary of the Invention

[0004] The purpose of this disclosure is to provide a parking trajectory planning method, storage medium, and vehicle that can achieve efficient and high-precision parking trajectory planning.

[0005] To achieve the above objectives, in a first aspect, this disclosure provides a parking trajectory planning method, comprising: determining multiple extended positions based on the relative positional relationship between the starting position and the parking position of the vehicle, wherein there is a drivable trajectory from the parking position to the extended positions; and determining a planned trajectory from the starting position to the parking position based on the multiple extended positions.

[0006] Optionally, determining multiple extended positions based on the relative positional relationship between the vehicle's starting position and parking position includes:

[0007] Multiple driving trajectories are determined based on the relative positional relationship and the parking position, and the driving trajectories conform to the kinematic equations of the vehicle.

[0008] The multiple driving trajectories are sampled to determine the multiple extended locations.

[0009] Optionally, determining multiple driving trajectories based on the relative positional relationship and the parking position includes:

[0010] The parking location type is determined based on the relative positional relationship;

[0011] Multiple driving trajectories are determined based on the parking location type and the parking location.

[0012] Optionally, determining multiple driving trajectories based on the parking location type and the parking location includes:

[0013] If the parking location type is a perpendicular parking space, the first extended location is determined based on the first preset driving direction, the first preset driving distance, and the parking location;

[0014] Multiple driving trajectories are generated based on multiple first preset attitude angles and the first extended position;

[0015] The perpendicular parking space is either a rear-entry perpendicular parking space or a front-entry perpendicular parking space, and the first preset driving direction corresponding to the rear-entry perpendicular parking space is opposite to the first preset driving direction corresponding to the front-entry perpendicular parking space.

[0016] Optionally, determining multiple driving trajectories based on the parking location type and the parking location includes:

[0017] If the parking location type is a parallel parking space, the second extended location is determined based on the second preset driving direction, the second preset driving distance, and the parking location;

[0018] The third extension position is determined based on the third preset driving direction, the third preset driving distance, and the second extension position;

[0019] Multiple driving trajectories are generated based on multiple second preset attitude angles and the third extended position;

[0020] The parallel parking space is either a first parallel parking space or a second parallel parking space. The parking direction of the vehicle corresponding to the first parallel parking space is opposite to that of the vehicle corresponding to the second parallel parking space. The third preset driving direction corresponding to the first parallel parking space is opposite to that of the third preset driving direction corresponding to the second parallel parking space.

[0021] Optionally, determining multiple driving trajectories based on the parking location type and the parking location includes:

[0022] If the parking location type is an angled parking space, the fourth extended location is determined based on the fourth preset driving direction and the parking location;

[0023] Multiple driving trajectories are generated based on multiple third preset attitude angles, fourth preset driving distances, and the fourth extended position;

[0024] The inclined train position is either a rear-entry inclined train position or a front-entry inclined train position, and the fourth preset driving direction corresponding to the rear-entry inclined train position is opposite to the fourth preset driving direction corresponding to the front-entry inclined train position.

[0025] Optionally, determining multiple extended positions based on the relative positional relationship between the vehicle's starting position and parking position includes:

[0026] Obtain obstacle information in the space where the vehicle is located;

[0027] The plurality of extended positions are determined based on the obstacle information and the relative positional relationship.

[0028] Optionally, determining multiple extended positions based on the relative positional relationship between the vehicle's starting position and parking position includes:

[0029] Obtain obstacle information in the space where the vehicle is located;

[0030] The plurality of extended positions are determined based on the obstacle information and the relative positional relationship.

[0031] Optionally, determining the planned trajectory from the starting position to the parking position based on the plurality of extended positions includes:

[0032] Determine a first planned trajectory from the starting position to the target extended position, wherein the target extended position is an extended position among the plurality of extended positions;

[0033] Determine a second planned trajectory from the target extension location to the parking location;

[0034] Based on the first planned trajectory and the second planned trajectory, the planned trajectory from the starting position to the parking position is determined.

[0035] Optionally, the parking trajectory planning method further includes:

[0036] A fast expanding random tree is generated according to the fast expanding random tree algorithm, wherein the root node of the fast expanding random tree is the starting position, and the ending growth condition of the fast expanding random tree includes the node of the fast expanding random tree reaching any of the expansion positions;

[0037] The determination of the first planned trajectory from the starting position to the target extended position includes:

[0038] The target expansion position is determined based on the distance between the nodes of the rapidly expanding random tree and each of the expansion positions;

[0039] Based on the rapidly expanding random tree, a first planned trajectory from the starting position to the target expansion position is determined.

[0040] Optionally, determining the first planned trajectory from the starting position to the target extended position includes:

[0041] Based on the fast expanding random tree and parent node search algorithm, the first planned trajectory is backtracked from the target expansion position to the starting position.

[0042] Optionally, determining the second planned trajectory from the target extended location to the parking location includes:

[0043] Determine the driving trajectory from the parking location to the target extended location;

[0044] Based on the opposite direction of the driving trajectory, a second planned trajectory is determined, starting from the target expansion position and ending at the parking position.

[0045] Optionally, the trajectory search strategy of the fast expanding random tree conforms to the kinematic equations of the vehicle.

[0046] Optionally, the trajectory search strategy includes:

[0047] If the trajectory endpoint includes attitude angle information, a trajectory from the trajectory start point to the trajectory endpoint is constructed based on the Dobbins curve; and / or,

[0048] If the trajectory endpoint does not include attitude angle information, the trajectory from the trajectory starting point to the trajectory endpoint is constructed based on the target attitude angle and the point chasing algorithm.

[0049] Optionally, the parking trajectory planning method further includes:

[0050] The target attitude angle is determined based on the distance between the trajectory start point and the trajectory end point, the attitude angle of the trajectory start point, the wheelbase of the vehicle, and the vector angle from the trajectory start point to the trajectory end point.

[0051] Optionally, the growth process of the rapidly expanding random tree includes:

[0052] Sampling points are obtained by taking samples in the space where the vehicle is located;

[0053] The target node is determined based on the distance between each node of the rapidly expanding random tree and the sampling point;

[0054] Based on the trajectory search strategy of the fast expanding random tree, construct the trajectory from the target node to the sampling point;

[0055] Based on the target node, the trajectory from the target node to the sampling point, and the preset travel distance, a new trajectory point is determined;

[0056] The new trajectory point is added as a new node to the fast-expanding random tree.

[0057] Optionally, the step of sampling in the space where the vehicle is located to obtain sampling points includes:

[0058] According to a preset sampling probability, sampling is performed in the space where the vehicle is located to obtain sampling points; the preset sampling probability is used to characterize the probability that the location corresponding to the sampling point belongs to the plurality of extended locations.

[0059] Optionally, the parking trajectory planning method further includes:

[0060] Obtain obstacle information in the space where the vehicle is located;

[0061] The step of adding the new trajectory point as a new node to the fast-expanding random tree includes:

[0062] Based on the obstacle information, determine whether the trajectory from the target node to the new trajectory point contains obstacles;

[0063] If the trajectory from the target node to the new trajectory point does not contain any obstacles, the new trajectory point is added as a new node to the fast-expanding random tree.

[0064] In a second aspect, this disclosure provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the steps of the parking trajectory planning method provided in the first aspect of this disclosure.

[0065] Thirdly, this disclosure provides a vehicle, including: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to: perform the steps of the parking trajectory planning method provided in the first aspect of this disclosure.

[0066] The above technical solution first determines multiple extended positions based on the relative positional relationship between the vehicle's starting position and parking position. Then, based on these extended positions, a planned trajectory from the starting position to the parking position is determined. On one hand, different relative positional relationships can represent different parking scenario types. The extended positions determined based on these relationships are adapted to the parking scenario, improving the accuracy of trajectory planning. On the other hand, a drivable trajectory exists between the parking position and the extended positions, allowing trajectory planning from the starting position to the parking position to be achieved based on the extended positions. Compared to directly planning the trajectory based on the parking position, this reduces the complexity of trajectory planning, increases efficiency, and correspondingly improves accuracy. Therefore, this technical solution enables efficient and high-precision parking trajectory planning.

[0067] Other features and advantages of this disclosure will be described in detail in the following detailed description section. Attached Figure Description

[0068] The accompanying drawings are provided to further illustrate the present disclosure and form part of the specification. They are used together with the following detailed description to explain the present disclosure, but do not constitute a limitation thereof. In the drawings:

[0069] Figure 1 This is a flowchart of a parking trajectory planning method according to an embodiment of the present disclosure.

[0070] Figures 2A to 2C This is an example diagram of a parking location type according to an embodiment of the present disclosure.

[0071] Figures 3A-3C This is an example diagram of an extended cluster according to an embodiment of the present disclosure.

[0072] Figure 4 This is a flowchart illustrating the generation of a fast random expansion tree according to an embodiment of the present disclosure.

[0073] Figure 5 This is a flowchart of a parking trajectory planning algorithm according to an embodiment of the present disclosure.

[0074] Figures 6A to 6C This is an example diagram of a parking trajectory constructed according to an embodiment of the present disclosure.

[0075] Figure 7 A structural block diagram of a parking trajectory planning method according to an embodiment of the present disclosure.

[0076] Figure 8 This is a structural block diagram of a vehicle according to an embodiment of the present disclosure. Detailed Implementation

[0077] The specific embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit this disclosure.

[0078] In this disclosure, unless otherwise stated, directional terms such as "up," "down," "left," "right," "front," and "back" are used only for the convenience of describing this disclosure and for simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this disclosure.

[0079] As mentioned in the background section, with the development of vehicle technology, vehicles can park using planned parking trajectories. The trajectory search algorithm determines the quality of the parking trajectory; the quality of the parking trajectory determines whether the vehicle can park successfully.

[0080] In automatic parking, the quality of trajectory planning directly affects whether a vehicle can successfully park or exit a parking space. On one hand, the vehicle's trajectory needs to avoid obstacles in the space to ensure the safety of passengers and the vehicle, and the trajectory must meet vehicle dynamics requirements so that the vehicle can accurately track the trajectory to the target point. The more numerous and widely distributed the obstacles in the parking environment, the narrower the drivable area, and the higher the difficulty of trajectory planning. Although search-based trajectory planning algorithms can generate collision-free vehicle trajectories, this initial trajectory usually does not meet vehicle dynamics requirements, and the vehicle cannot accurately track it. On the other hand, the efficiency of trajectory planning directly affects the user experience of the parking function. Because the trajectory search algorithm needs to start from the vehicle's starting position, try different driving directions and routes, and gradually explore the entire spatial area until reaching the vehicle's destination, the computational complexity of the search algorithm is usually high. The limited computing resources of the onboard chip cannot provide sufficient computing power for the search algorithm, resulting in high search time, significantly increasing user waiting time and reducing user satisfaction.

[0081] In related technologies, the parking trajectories generated by the trajectory search algorithms are of poor quality, resulting in the planned parking trajectories failing to meet user needs.

[0082] For example, some trajectory search algorithms use the target parking location as the starting point and the vehicle's initial position as the ending point, employing an improved hybrid algorithm to search for the trajectory on a rasterized map. However, due to the low quality of the obtained trajectory, further optimization is needed. Other trajectory search algorithms use a random tree search algorithm, starting with the vehicle's initial position and gradually expanding the tree until the parking location is reached, at which point the trajectory search ends. This approach uses a single target node as the final search node, resulting in very low search efficiency.

[0083] It is evident that the parking trajectory planning methods employed by the relevant technologies cannot guarantee the quality of the parking trajectory; or the planning efficiency of the parking trajectory is low.

[0084] Based on this, this disclosure provides a parking trajectory planning scheme. In this technical solution, different parking scenario types are considered, and the parking location (parking position) is expanded. Based on the expanded parking location, the parking trajectory is then planned. Compared with the trajectory planning based directly on a single parking location, the complexity is reduced. Thus, this technical solution can not only adapt to different parking scenarios, but also reduce the complexity of trajectory planning and improve the efficiency of trajectory planning.

[0085] Figure 1This is a flowchart illustrating a parking trajectory planning method according to an exemplary embodiment. This method can be applied to electronic devices with processing capabilities, such as in-vehicle processors and controllers. Figure 1 As shown, the method may include the following steps.

[0086] Step S11: Determine multiple extended positions based on the relative positional relationship between the vehicle's starting position and parking position.

[0087] In step S11, the vehicle's starting position can be understood as the vehicle's current position, and the vehicle's parking position can be understood as the target parking position, such as the location of the parking space.

[0088] Relative positional relationship can be understood as the positional relationship between the starting position and the parking position in the space where the vehicle is located. It can represent the position of the vehicle, the parking position and the positional relationship between the two. It can be understood that different relative positional relationships can correspond to different parking scenario types.

[0089] Examples include starting position directly in front of the parking space, starting position to the left front of the parking space, starting position behind the parking space, starting position to the right rear of the parking space, etc.

[0090] In step S11, there is a drivable trajectory from the parking position to the extended position. This can be understood as the vehicle being able to drive from the parking position to the extended position. Therefore, the vehicle can also drive from the extended position to the parking position in the opposite direction of the drivable trajectory. Thus, the extended position can be used to plan the parking trajectory from the starting position to the parking position.

[0091] In some embodiments, based on the premise that there is a drivable trajectory from the parking location to the extended location, multiple trajectories starting from the parking location can be constructed first, and then the extended location can be determined based on these multiple trajectories.

[0092] Therefore, as an optional implementation, step S11 includes: determining multiple driving trajectories based on relative positional relationships and parking positions, wherein the driving trajectories conform to the kinematic equations of the vehicle; sampling the multiple driving trajectories to determine multiple extended positions.

[0093] In this implementation, starting from the parking position, multiple driving trajectories can be constructed based on different relative positional relationships using the vehicle's kinematic equations.

[0094] The kinematic equations of a vehicle can be used to characterize the vehicle's driving mode and can be understood as the vehicle's dynamic constraints. If the vehicle's kinematic equations are not considered during trajectory planning, the planned trajectory may not meet the vehicle's kinematic requirements and cannot be called a drivable trajectory.

[0095] For example, suppose there are two points, I and J, in space. Point J is to the right and in front of point I, and there is an obstacle in front of point I. Then, for a vehicle to travel from point I to point J, it cannot travel directly along a straight line from point I to point J. Instead, it must travel along an arc-shaped trajectory from point I to point J, while avoiding the obstacle.

[0096] As an optional implementation, multiple driving trajectories are determined based on relative positional relationships and parking locations, including: determining parking location types based on relative positional relationships; and determining multiple driving trajectories based on parking location types and parking locations.

[0097] In this implementation, different relative positional relationships can correspond to different parking location types.

[0098] This disclosure may involve three types of parking spaces: perpendicular parking spaces, parallel parking spaces, and angled parking spaces.

[0099] For example, please refer to Figures 2A to 2C This is an example diagram illustrating the three parking location types provided in this disclosure. Figure 2A In the diagram, parking spaces S1 are vertically distributed in the space where the vehicle is located. The vehicle may be located at positions A1, A2, and A3, or other possible positions.

[0100] exist Figure 2B In the parking space S2, the parking spaces are arranged in parallel within the space where the vehicles are located. The vehicles may be located in positions B1, B2, and B3, or in other possible positions.

[0101] exist Figure 2C In the parking space S3, the parking space is arranged diagonally in the space where the vehicle is located. The vehicle may be located in positions C1, C2, and C3, or more possible positions.

[0102] Based on the example diagrams of these three parking location types, it can be seen that when the relative positional relationship is known, the parking location type can be determined. Furthermore, different driving trajectory generation methods can be adopted for different parking location types.

[0103] As a first optional implementation, if the parking location type is a perpendicular parking space, a first extended position is determined based on a first preset driving direction, a first preset driving distance, and the parking location; multiple driving trajectories are generated based on multiple first preset attitude angles and the first extended position; wherein, the perpendicular parking space is a rear-end perpendicular parking space or a front-end perpendicular parking space, and the first preset driving direction corresponding to the rear-end perpendicular parking space is opposite to the first preset driving direction corresponding to the front-end perpendicular parking space.

[0104] In this embodiment, the first extended position can be the position reached by the vehicle after traveling a first preset distance along a preset driving direction from the parking position.

[0105] Multiple first preset attitude angles are different attitude angles, and one first preset attitude angle can correspond to one driving trajectory.

[0106] The generation of multiple driving trajectories includes: the vehicle driving from the first extended position using different first preset attitude angles, and the corresponding driving trajectories are the generated multiple driving trajectories.

[0107] Rear-entry perpendicular parking spaces mean that the rear of the vehicle enters the parking space first. Front-entry perpendicular parking spaces mean that the front of the vehicle enters the parking space first. Therefore, the first preset driving directions for front-entry and rear-entry perpendicular parking spaces are opposite.

[0108] As a second optional implementation, if the parking location type is a parallel parking space, a second extended position is determined based on a second preset driving direction, a second preset driving distance, and the parking location; a third extended position is determined based on a third preset driving direction, a third preset driving distance, and the second extended position; multiple driving trajectories are generated based on multiple second preset attitude angles and the third extended position; wherein, the parallel parking space is either a first parallel parking space or a second parallel parking space, the vehicle exit direction corresponding to the first parallel parking space is opposite to the vehicle exit direction corresponding to the second parallel parking space, and the third preset driving direction corresponding to the first parallel parking space is opposite to the third preset driving direction corresponding to the second parallel parking space.

[0109] In this embodiment, the second extended position can be the position reached by the vehicle after traveling a second preset distance along the second preset driving direction, starting from the parking position.

[0110] The third extended position can be the position reached by the vehicle after continuing to travel from the second extended position and traveling a third preset distance along the third preset driving direction.

[0111] Multiple second preset attitude angles, each with a different attitude angle, can correspond to a driving trajectory.

[0112] The generation of multiple driving trajectories includes: the vehicle driving from the third extended position using different second preset attitude angles, and the corresponding driving trajectories are the generated multiple driving trajectories.

[0113] For example, the first parallel parking space can be a left-exit parallel parking space, meaning that the vehicle drives out of the parking space from the left. The second parallel parking space can be a right-exit parallel parking space, meaning that the vehicle drives out of the parking space from the right. Therefore, the third preset driving direction corresponding to the first parallel parking space is opposite to the third preset driving direction corresponding to the second parallel parking space.

[0114] As a third optional implementation, if the parking position type is an angled parking space, a fourth extended position is determined based on the fourth preset driving direction, the fourth preset driving distance, and the parking position; multiple driving trajectories are generated based on multiple third preset attitude angles and the fourth extended position; wherein, the angled parking space is either a rear-entry angled parking space or a front-entry angled parking space, and the fourth preset driving direction corresponding to the rear-entry angled parking space is opposite to the fourth preset driving direction corresponding to the front-entry angled parking space.

[0115] The fourth extended position can be the position reached by the vehicle starting from the parking position and traveling a fourth preset distance along the fourth preset driving direction.

[0116] Multiple third preset attitude angles are different attitude angles, and one third preset attitude angle can correspond to one driving trajectory.

[0117] The generation of multiple driving trajectories includes: the vehicle driving from the fourth extended position using different third preset attitude angles, and the corresponding driving trajectories are the generated multiple driving trajectories.

[0118] A rear-entry angled parking space means that the rear of the vehicle enters the parking space first; a front-entry angled parking space means that the front of the vehicle enters the parking space first. Therefore, the fourth preset driving direction corresponding to a rear-entry angled parking space is opposite to the fourth preset driving direction corresponding to a front-entry angled parking space.

[0119] In the above three embodiments, the first preset attitude angle, the second preset attitude angle, and the third preset attitude angle can be preset according to the specific parking scenario; and the first preset driving distance, the second preset driving distance, the third preset driving distance, and the fourth preset driving distance can also be preset according to the specific parking scenario.

[0120] Regardless of which implementation method is used, after generating multiple driving trajectories, sampling these multiple driving trajectories can determine multiple extended locations.

[0121] In some embodiments, the parking location can be regarded as the target point in the space where the vehicle is located, and multiple extended locations can be regarded as multiple points in the space where the vehicle is located, which can be called extended points. These multiple extended points can form an extended cluster, and the process of generating the extended cluster is the process of extending the target point.

[0122] Please refer to the following. Figures 3A-3B Here is an example diagram of the extended clusters corresponding to the three parking location types, where... Figure 3A This is an example diagram of the extended cluster corresponding to a vertical parking space. Figure 3B This is an example diagram of the extended clusters corresponding to parallel parking spaces. Figure 3C This is an example diagram of the extended cluster corresponding to the slanted train position.

[0123] Combination Figure 3A The target point expansion method for rear-end parking perpendicular spaces can be as follows: Assuming the vehicle is already at target point S, first, the vehicle moves forward a certain distance to reach point A. Then, it continues to move forward using different attitude angles, generating multiple trajectory clusters. These trajectory clusters are sampled to obtain the final target point expansion cluster.

[0124] Continue to combine Figure 3A The target point expansion method for a front-end parking perpendicular parking space can be as follows: Assuming the vehicle is already at target point S, the vehicle first reverses a certain distance to point A. Then, it continues to reverse using different attitude angles, generating multiple trajectory clusters. These multiple trajectories are sampled to obtain the final target point expansion cluster.

[0125] Combination Figure 3B The target point expansion method for left-exit parallel parking spaces can be as follows: Assuming the vehicle is already at target point S, the vehicle first reverses a certain distance to reach point A. Then, it turns the steering wheel fully to the left and moves forward a certain distance to reach point B. Finally, it continues to move forward using different attitude angles, generating multiple trajectory clusters. Sampling these multiple trajectories yields the final target point expansion cluster.

[0126] Continue to combine Figure 3B The target point expansion method for right-exit parallel parking spaces can be as follows: Assuming the vehicle is already at target point S, the vehicle first reverses a certain distance to reach point A. Then, it turns the steering wheel fully to the right and moves forward a certain distance to reach point B. Finally, it continues to move forward using different attitude angles, generating multiple trajectory clusters. Sampling these multiple trajectories yields the final target point expansion cluster.

[0127] Combination Figure 3C The target point expansion method for rear-end parking at an inclined train position can be as follows: Assuming the vehicle is already at target point S, the vehicle first moves a certain distance to point A. Then, it continues to move forward using different attitude angles, generating multiple trajectory clusters. These multiple trajectories are sampled to obtain the final target point expansion cluster.

[0128] Continue to combine Figure 3CThe target point expansion method for a train-head-parking inclined parking space can be as follows: Assuming the vehicle is already at target point S, the vehicle first reverses a certain distance to reach point A, and then continues to reverse using different attitude angles, generating multiple trajectory clusters. Sampling is performed on these multiple trajectories to obtain the final target point expansion cluster.

[0129] It is understandable that the parameters such as straight-line distance, back-up distance, and attitude angle in the above examples can be adjusted according to the actual scenario.

[0130] Furthermore, in this disclosure, each point in the extended cluster contains the spatial position coordinates and attitude angles of the vehicle's location in space.

[0131] In some embodiments, if there are no obstacles in the space where the vehicle is located, then the determination of multiple extended locations does not need to take obstacles into account.

[0132] In other embodiments, if there are obstacles in the space where the vehicle is located, the determination of multiple extension locations needs to take the obstacles into account. Therefore, as an optional implementation, step S11 includes: obtaining obstacle information in the space where the vehicle is located; and determining multiple extension locations based on the obstacle information and relative positional relationships.

[0133] In the implementation of step S11 above, when sampling multiple driving trajectories, multiple extended positions can be determined based on obstacle information.

[0134] For example, based on obstacle information, it is determined whether there is an obstacle between the currently sampled position and the parking position. If there is an obstacle, the currently sampled position is not the extended position; if there is no obstacle, the currently sampled position is the extended position.

[0135] Step S12: Determine the planned trajectory from the starting position to the parking position based on multiple extended positions.

[0136] In step S12, multiple extended positions can be regarded as multiple extended points, and multiple extended points can form an extended cluster. In order to realize the planned trajectory from the starting position to the parking position, any point can be found from the extended cluster first, and then the final parking trajectory can be constructed using this point, the starting position and the parking position.

[0137] Therefore, as an optional implementation, step S12 includes: determining a first planned trajectory from the starting position to the target extended position, wherein the target extended position is an extended position among a plurality of extended positions; determining a second planned trajectory from the target extended position to the parking position; and determining a planned trajectory from the starting position to the parking position based on the first planned trajectory and the second planned trajectory.

[0138] In this implementation, the target expansion position is an expansion position among multiple expansion positions. Therefore, it is necessary to first find the target expansion position among multiple expansion positions.

[0139] In some embodiments, the starting position can be expanded using a fast expanding random tree algorithm until the fast expanding random tree reaches any point in the expansion cluster formed by multiple expansion positions, at which point the target expansion position is considered to have been found.

[0140] Therefore, as an optional implementation, the parking trajectory planning method further includes: generating a fast expanding random tree according to the fast expanding random tree algorithm, wherein the root node of the fast expanding random tree is the starting position, and the termination growth condition of the fast expanding random tree includes the node of the fast expanding random tree reaching any expansion position.

[0141] Then, determining the first planned trajectory from the starting position to the target expansion position includes: determining the target expansion position based on the distance between the nodes of the fast expanding random tree and each expansion position; and determining the first planned trajectory from the starting position to the target expansion position based on the fast expanding random tree.

[0142] Rapidly Exploring Random Trees (RRT) are a type of random tree that expands rapidly.

[0143] In some embodiments, the trajectory search strategy of the fast expanding random tree conforms to the vehicle's kinematic equations. Therefore, during trajectory search, the trajectory needs to be searched according to the vehicle's kinematic equations; for example, the searched trajectory may be curved.

[0144] To conform to the vehicle's kinematic equations, the trajectory search strategy may include: if the trajectory endpoint includes attitude angle information, constructing a trajectory from the trajectory start point to the trajectory endpoint based on the Dubins curve; and / or, if the trajectory endpoint does not include attitude angle information, constructing a trajectory from the trajectory start point to the trajectory endpoint based on the target attitude angle and a point chasing algorithm.

[0145] It is understood that, based on the description of the foregoing embodiments, if the trajectory endpoint belongs to a point in the extended cluster, the point in the extended cluster has attitude angle information; while if the trajectory endpoint does not belong to a point in the extended cluster, it does not have attitude angle information.

[0146] As an optional implementation, the target attitude angle is determined based on the distance between the trajectory start point and the trajectory end point, the attitude angle of the trajectory start point, the wheelbase of the vehicle, and the vector angle from the trajectory start point to the trajectory end point.

[0147] In this implementation, the vehicle's wheelbase is known information and can be directly obtained; the vector angle can be understood as the angle between the vector from the trajectory start point to the trajectory end point and the x-axis; the trajectory start point is a node searched in the fast expanding random tree, which has attitude angle information. The distance between the trajectory start point and the trajectory end point can be calculated and determined based on the spatial coordinates of the two points.

[0148] For example, the formula for calculating the target attitude angle can be: θ1 represents the angle between the vector from the starting point to the ending point of the trajectory and the x-axis, θ2 represents the attitude angle of the starting point of the trajectory, θ represents the wheelbase of the vehicle, and s represents the distance between the starting point and the ending point of the trajectory.

[0149] In some embodiments, the process of generating a rapidly expanding random tree may include an initialization phase, a growth phase, and an end phase.

[0150] As an optional implementation, the growth process of the fast-expanding random tree may include: sampling in the space where the vehicle is located to obtain sampling points; determining a target node based on the distance between each node of the fast-expanding random tree and the sampling point; constructing a trajectory from the target node to the sampling point according to the trajectory search strategy of the fast-expanding random tree; determining a new trajectory point based on the target node, the trajectory from the target node to the sampling point, and a preset driving distance; and adding the new trajectory point as a new node to the fast-expanding random tree.

[0151] In this implementation, if the requirement is to further improve the efficiency of trajectory planning, obtaining sampling points includes: sampling in the space where the vehicle is located according to a preset sampling probability to obtain sampling points; the preset sampling probability is used to characterize the probability that the location corresponding to the sampling point belongs to multiple extended locations, that is, the probability that the sampling point belongs to an extended cluster.

[0152] For example, the preset sampling probability can be 50%, which means that the probability of the location corresponding to the sampling point belonging to multiple extensions is 50%.

[0153] Correspondingly, the probability that the location corresponding to the sampling point does not belong to multiple extended locations is: 1 - preset sampling probability.

[0154] In some embodiments, the distance between each node and the sampling point can be Euclidean distance. When calculating this distance, the planar distance can be calculated, without needing to calculate the spatial distance.

[0155] In some embodiments, the target node is the node that is closest to the sampling point.

[0156] In some embodiments, if the sampling point belongs to the extended cluster, the trajectory from the target node to the sampling point is constructed based on the Dubins curve; if the sampling point does not belong to the extended cluster, a point chasing method is used to move forward from the target node at a fixed attitude angle to reach the sampling point, thereby constructing the trajectory.

[0157] In some embodiments, starting from the target node, the vehicle travels forward a preset distance along the trajectory from the target node to the sampling point, and the reached trajectory point is the new trajectory point.

[0158] It's understandable that if there are no obstacles in the space where the vehicle is located, then obstacles don't need to be considered during the generation of the fast expanding random tree. However, if obstacles exist in the space where the vehicle is located, then obstacles need to be considered during the generation of the fast expanding random tree.

[0159] Therefore, as an optional implementation, the parking trajectory planning method further includes: obtaining obstacle information in the space where the vehicle is located; adding the new trajectory point as a new node to the fast expanding random tree, including: determining whether the trajectory from the target node to the new trajectory point contains obstacles based on the obstacle information; if the trajectory from the target node to the new trajectory point does not contain obstacles, adding the new trajectory point as a new node to the fast expanding random tree.

[0160] In this way, it can be ensured that the planned trajectory does not contain obstacles, resulting in higher accuracy of the final planned trajectory.

[0161] For example, please refer to Figure 4 This is a flowchart illustrating the construction process of the Rapidly Expanding Random Tree (hereinafter referred to as RRT tree) provided in this disclosure. An expansion cluster refers to a cluster of points formed by expansion points corresponding to multiple expansion positions. The construction process of the RRT tree includes:

[0162] During the initialization phase, the starting point (i.e., the initial position) is used as the root node to initialize the RRT tree. At this time, the RRT tree only contains the starting point node.

[0163] Generation phase:

[0164] Step S41: Randomly sample point P.

[0165] In step S41, a random sampling function can be used to sample point P from space. The random sampling function has a 50% probability of selecting a point P from the extended cluster; and a 50% probability of randomly generating a point P in space. If point P is selected from the extended cluster, it contains both spatial position coordinates and attitude angles; if point P is randomly generated, it only contains the left side of the spatial position coordinates and does not contain the attitude angles.

[0166] Step S42: Select the node N in the RRT tree that is closest to point P.

[0167] In step S42, the distance can be calculated using the following formula: P x and P y N represents the spatial coordinates of point P. x and N y This represents the spatial coordinates of point N. The distance calculation formula does not consider the attitude angle.

[0168] Step S43: Based on the type of point P, use the Dubins curve or point chasing method to establish the trajectory from node N to point P.

[0169] In step S43, after point P is randomly selected from the target point expansion cluster in step S41, since point P itself contains attitude angle information, the trajectory from point N to point P can be constructed using the Dubins curve.

[0170] After point P is randomly generated from space in step S41, since point P does not contain attitude angle information, a point-chasing method is needed to move from point N with a fixed attitude angle to reach point P. This trajectory does not consider the attitude angle at the final point P, which can quickly expand the RRT tree from point N to point P, thereby improving the RRT expansion efficiency.

[0171] The formula for calculating a fixed attitude angle can be: θ1 represents the angle between the vector from point N to point P and the x-axis, θ2 represents the attitude angle of point N, L represents the vehicle wheelbase, and s represents the distance between point N and point P.

[0172] It's understandable that using the Dobbins curve and point chasing methods to expand the RRT tree can effectively improve its expansion efficiency. If N points in the RRT tree are surrounded by obstacles, the point chasing method can help the vehicle escape the predicament as quickly as possible. The Dobbins curve, on the other hand, considers the vehicle's attitude angles to smoothly connect the RRT tree to the target point expansion cluster.

[0173] Step S44: Start from point N and travel a certain distance along the trajectory to point A. If the trajectory from point N to point A does not contain any obstacles, add point A as a new node to the RRT tree; return to step S41 and continue the RRT tree generation process.

[0174] In the final stage, the RRT tree ends its growth when a node of the RRT tree reaches any point in the extended cluster, or when the maximum number of growth iterations is reached.

[0175] It is understandable that if the trajectory from point N to point A passes through point P (a point in the extended cluster), it means that the RRT tree can reach a certain point in the extended cluster from the starting point, and at this time the RRT tree is completed.

[0176] Based on the above introduction to the fast random expansion tree, assuming the target expansion position is considered as the target point in the expansion cluster, it can be understood that the target expansion point can be any intersection of the fast random expansion tree and the expansion cluster.

[0177] In some embodiments, after the fast-expanding random tree finishes growing, there may be more than one intersection point between the fast-expanding random tree and the expanding cluster. In this case, the distance between each intersection point and any node of the fast-expanding random tree can be calculated, and the intersection point with the smallest distance can be selected; or an intersection point with a distance less than a preset distance can be selected, where the preset distance can be 0. Furthermore, the distance calculated in this embodiment is a spatial distance.

[0178] For example, the formula for calculating distance can be: Where T represents any point in the extended cluster and R represents any point in the RRT tree, this distance calculation formula is used to calculate the distance based on the spatial position and attitude angle.

[0179] It is understandable that if there exist two points M and N, where M belongs to the extended cluster and N belongs to the RRT tree, and their distance is less than a preset distance, then it means that the RRT tree and the extended cluster have been successfully connected, and a trajectory from the starting position to the parking position can be constructed by trajectory backtracking. Otherwise, the trajectory search fails.

[0180] As an optional implementation, determining the first planned trajectory from the starting position to the target expansion position includes: backtracking the first planned trajectory from the target expansion position to the starting position according to a fast expanding random tree and a parent node search algorithm.

[0181] For example, starting from the target expansion position, the parent node is searched step by step in the rapidly expanding random tree until the starting position is backtracked, thus constructing the first planned trajectory.

[0182] As an optional implementation, determining a second planned trajectory from the target expansion location to the parking location includes: determining a driving trajectory from the parking location to the target expansion location; and determining a second planned trajectory starting from the target expansion location and ending at the parking location, based on the opposite direction of the driving trajectory.

[0183] For example, starting from the target expansion location, the trajectory is traced back to the parking location in the opposite direction of the generated driving trajectory at the target expansion location to construct a second planned trajectory.

[0184] Since both the extended cluster and the RRT tree are determined based on the vehicle's kinematics, the first and second planned trajectories do not require further optimization.

[0185] In some embodiments, the first planned trajectory and the second planned trajectory are merged to determine the final parking trajectory.

[0186] In some embodiments, if the connection point between the first and second planned trajectories is completely consistent, the merged parking trajectory does not require optimization. If there is a deviation between the connection points of the first and second planned trajectories, smoothing techniques can be used to smooth the connection points of the two planned trajectories to optimize the parking trajectory.

[0187] For example, smoothing methods can include polynomial curve smoothing, spline curve smoothing, etc.

[0188] Please refer to the following. Figure 5 The flowchart below shows the parking trajectory planning algorithm provided in this disclosure. The planning process for this parking trajectory planning includes:

[0189] Step S51: Expand the target point to generate an expanded cluster.

[0190] In step S51, the extended cluster is understood as a cluster composed of multiple extended positions, and one extended position corresponds to one extended point in the extended cluster; the target point is the parking position.

[0191] Step S52: Expand the starting point and construct a fast-expanding random tree.

[0192] In step S52, the starting point is understood as the starting position.

[0193] Step S53, trajectory backtracking.

[0194] Step S54: Smooth the connection points.

[0195] It is understood that the specific implementation methods of steps S51 to S54 are the same as those described in the previous embodiments, and will not be repeated here.

[0196] Please refer to the following. Figures 6A to 6C This is an example diagram showing the parking trajectories constructed under the three parking scenarios provided in this disclosure. Figures 6A to 6C In the diagram, point Q represents the starting position, point J represents the connection point between the first and second planned trajectories, and point S represents the parking position. It can be understood that the trajectory from point Q to point J is the first planned trajectory, and the trajectory from point J to point S is the second planned trajectory.

[0197] Figure 6A The diagram shows the parking trajectory constructed in a vertical parking scenario. path1 represents the first planned trajectory in the scenario, path2 represents the second planned trajectory in the scenario, RRT1 represents the RRT tree constructed in the scenario, K1 represents the extended cluster in the scenario, and O1, O2 and O3 represent obstacles at different locations.

[0198] Figure 6BThe diagram shows the parking trajectory constructed in a parallel parking scenario. path3 represents the first planned trajectory, path4 represents the second planned trajectory in the scenario, RRT2 represents the RRT tree constructed in the scenario, K2 represents the extended cluster in the scenario, and O4, O5, and O6 represent obstacles at different locations.

[0199] Figure 6C The diagram shows the parking trajectory constructed in the diagonal parking scenario. path5 represents the first planned trajectory, path6 represents the second planned trajectory in the scenario, RRT3 represents the RRT tree constructed in the scenario, K3 represents the extended cluster in the scenario, and O7, O8 and O9 represent obstacles at different locations.

[0200] As can be seen from the embodiments described in this disclosure, on the one hand, this disclosure expands the target point according to different parking scenario types, using vehicle kinematic equations and spatial obstacle information, to obtain an expanded cluster. When the search algorithm reaches any point in the expanded cluster, it can enter the final target point location along the opposite direction of the expansion direction, greatly expanding the number of target points searched and effectively improving search efficiency.

[0201] On the other hand, this disclosure employs a point-chasing + Durbins curve-based RRT trajectory search algorithm to quickly construct a drivable trajectory from the starting point to any point in the extended cluster. While satisfying the probabilistic completeness of the search process, the algorithm also obtains a trajectory that meets the vehicle's kinematic requirements, allowing the vehicle to easily follow the trajectory to the final target point and smoothly complete the entire parking process.

[0202] Therefore, this technical solution can achieve efficient and high-precision parking trajectory planning.

[0203] Based on the same concept, this disclosure also provides a parking trajectory planning device. Figure 7 This is a block diagram illustrating a parking trajectory planning device according to an exemplary embodiment. Figure 7 As shown, the parking trajectory planning device 700 may include:

[0204] The extension module 701 is configured to: determine multiple extension positions based on the relative positional relationship between the vehicle's starting position and parking position, wherein there is a drivable trajectory from the parking position to the extension position;

[0205] The planning module 702 is configured to: determine the planned trajectory from the starting position to the parking position based on the plurality of extended positions.

[0206] Optionally, the expansion module 701 includes:

[0207] The first trajectory determination module is configured to determine multiple driving trajectories based on the relative positional relationship and the parking position, wherein the driving trajectories conform to the kinematic equations of the vehicle.

[0208] The sampling module is configured to sample the multiple driving trajectories and determine the multiple extended positions.

[0209] Optionally, the first trajectory determination module is further configured to: determine the parking location type based on the relative positional relationship; and determine multiple driving trajectories based on the parking location type and the parking location.

[0210] Optionally, the first trajectory determination module is further configured to:

[0211] If the parking location type is a perpendicular parking space, the first extended location is determined based on the first preset driving direction, the first preset driving distance, and the parking location;

[0212] Multiple driving trajectories are generated based on multiple first preset attitude angles and the first extended position;

[0213] The perpendicular parking space is either a rear-entry perpendicular parking space or a front-entry perpendicular parking space, and the first preset driving direction corresponding to the rear-entry perpendicular parking space is opposite to the first preset driving direction corresponding to the front-entry perpendicular parking space.

[0214] Optionally, the first trajectory determination module is further configured to:

[0215] If the parking location type is a parallel parking space, the second extended location is determined based on the second preset driving direction, the second preset driving distance, and the parking location;

[0216] The third extension position is determined based on the third preset driving direction, the third preset driving distance, and the second extension position;

[0217] Multiple driving trajectories are generated based on multiple second preset attitude angles and the third extended position;

[0218] The parallel parking space is either a first parallel parking space or a second parallel parking space. The parking direction of the vehicle corresponding to the first parallel parking space is opposite to that of the vehicle corresponding to the second parallel parking space. The third preset driving direction corresponding to the first parallel parking space is opposite to that of the third preset driving direction corresponding to the second parallel parking space.

[0219] Optionally, the first trajectory determination module is further configured to: if the parking position type is an inclined train parking space, determine a fourth extended position based on a fourth preset driving direction, a fourth preset driving distance and the parking position;

[0220] Multiple driving trajectories are generated based on multiple third preset attitude angles and the fourth extended position;

[0221] The inclined train position is either a rear-entry inclined train position or a front-entry inclined train position, and the fourth preset driving direction corresponding to the rear-entry inclined train position is opposite to the fourth preset driving direction corresponding to the front-entry inclined train position.

[0222] Optionally, the expansion module 701 includes:

[0223] The first acquisition module is configured to acquire obstacle information in the space where the vehicle is located;

[0224] An extended position determination module is configured to determine the plurality of extended positions based on the obstacle information and the relative positional relationship.

[0225] Optionally, planning module 702 includes:

[0226] The second trajectory determination module is configured to determine a first planned trajectory from the starting position to the target extended position, wherein the target extended position is an extended position among the plurality of extended positions;

[0227] The third trajectory determination module is configured to determine a second planned trajectory from the target extension location to the parking location;

[0228] The fourth trajectory determination module is configured to determine the planned trajectory from the starting position to the parking position based on the first planned trajectory and the second planned trajectory.

[0229] Optionally, the parking trajectory planning device 700 also includes:

[0230] The generation module is configured to generate a fast expanding random tree according to the fast expanding random tree algorithm, wherein the root node of the fast expanding random tree is the starting position, and the ending growth condition of the fast expanding random tree includes the node of the fast expanding random tree reaching any of the expansion positions.

[0231] The second trajectory determination module is further configured to: determine the target expansion position based on the distance between the nodes of the fast-expanding random tree and each of the expansion positions;

[0232] Based on the rapidly expanding random tree, a first planned trajectory from the starting position to the target expansion position is determined.

[0233] Optionally, the second trajectory determination module is further configured to: backtrack the first planned trajectory from the target expansion position to the starting position according to the fast expanding random tree and parent node search algorithm.

[0234] Optionally, the second trajectory determination module is further configured to: determine the driving trajectory from the parking position to the target extended position;

[0235] Based on the opposite direction of the driving trajectory, a second planned trajectory is determined, starting from the target expansion position and ending at the parking position.

[0236] Optionally, the trajectory search strategy of the fast expanding random tree conforms to the kinematic equations of the vehicle.

[0237] Optionally, the trajectory search strategy includes:

[0238] If the trajectory endpoint includes attitude angle information, a trajectory from the trajectory start point to the trajectory endpoint is constructed based on the Durbins curve; and / or,

[0239] If the trajectory endpoint does not include attitude angle information, the trajectory from the trajectory starting point to the trajectory endpoint is constructed based on the target attitude angle and the point chasing algorithm.

[0240] Optionally, the parking trajectory planning device 700 further includes an attitude angle determination module, configured to determine the target attitude angle based on the distance between the trajectory start point and the trajectory end point, the attitude angle of the trajectory start point, the wheelbase of the vehicle, and the vector angle from the trajectory start point to the trajectory end point.

[0241] Optionally, the generation module is further configured as follows:

[0242] Sampling points are obtained by taking samples in the space where the vehicle is located;

[0243] The target node is determined based on the distance between each node of the rapidly expanding random tree and the sampling point;

[0244] Based on the trajectory search strategy of the fast expanding random tree, construct the trajectory from the target node to the sampling point;

[0245] Based on the target node, the trajectory from the target node to the sampling point, and the preset travel distance, a new trajectory point is determined;

[0246] The new trajectory point is added as a new node to the fast-expanding random tree.

[0247] Optionally, the generation module is further configured to: sample in the space where the vehicle is located according to a preset sampling probability to obtain sampling points; the preset sampling probability is used to characterize the probability that the location corresponding to the sampling point belongs to the plurality of extended locations.

[0248] Optionally, the acquisition module is further configured to: acquire obstacle information in the space where the vehicle is located;

[0249] The generation module is further configured as follows:

[0250] Based on the obstacle information, determine whether the trajectory from the target node to the new trajectory point contains obstacles;

[0251] If the trajectory from the target node to the new trajectory point does not contain any obstacles, the new trajectory point is added as a new node to the fast-expanding random tree.

[0252] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, which can be executed by the parking trajectory planning device 700 to complete the parking trajectory planning method described above. For example, the non-transitory computer-readable storage medium may be a ROM, a first random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0253] Please refer to Figure 8 , Figure 8 This is a functional block diagram of a vehicle according to an exemplary embodiment. The vehicle 800 may include various subsystems, such as an infotainment system 810, a perception system 820, a decision control system 830, a drive system 840, and a computing platform 850. The vehicle 800 may also include more or fewer subsystems, and each subsystem may include multiple components. Furthermore, each subsystem and each component of the vehicle 800 can be interconnected via wired or wireless means.

[0254] In some embodiments, the infotainment system 810 may include a communication system, an entertainment system, and a navigation system, etc.

[0255] The perception system 820 may include several sensors for sensing information about the environment surrounding the vehicle 800. For example, the perception system 820 may include a global positioning system (which may be GPS, BeiDou, or other positioning systems), an inertial measurement unit (IMU), lidar, millimeter-wave radar, ultrasonic radar, and a camera device.

[0256] The decision control system 830 may include a computing system, a vehicle controller, a steering system, a throttle, and a braking system.

[0257] The drive system 840 may include components that provide powered motion to the vehicle 800. In one embodiment, the drive system 840 may include an engine, an energy source, a transmission system, and wheels. The engine may be one or a combination of internal combustion engines, electric motors, and compressed air engines. The engine is capable of converting energy provided by the energy source into mechanical energy.

[0258] Some or all of the functions of the vehicle 800 are controlled by a computing platform 850. The computing platform 850 may include at least one processor 851 and a memory 852, the processor 851 being able to execute instructions 853 stored in the memory 852.

[0259] The processor 851 can be any conventional processor, such as a commercially available CPU. The processor may also include, for example, a Graphics Processing Unit (GPU), a Field Programmable Gate Array (FPGA), a System on Chip (SOC), an Application Specific Integrated Circuit (ASIC), or a combination thereof.

[0260] The memory 852 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0261] In addition to instruction set 853, memory 852 can also store data, such as road maps, route information, vehicle position, direction, speed, and other data. The data stored in memory 852 can be used by computing platform 850.

[0262] In this embodiment of the disclosure, processor 851 may execute instruction 853 to complete all or part of the steps of the parking trajectory planning method described above.

[0263] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0264] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0265] Furthermore, various different embodiments of this disclosure can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content disclosed in this disclosure.

Claims

1. A parking trajectory planning method, characterized in that, include: Based on the relative positional relationship between the vehicle's starting position and its parking position, multiple extended positions are determined, and there is a drivable trajectory from the parking position to the extended position; A fast expanding random tree algorithm is used to generate a fast expanding random tree, where the root node is the starting position and the termination condition includes the node of the fast expanding random tree reaching any of the expansion positions. The generation process of the fast expanding random tree includes: sampling in the space where the vehicle is located to obtain sampling points; determining target nodes based on the distances between each node of the fast expanding random tree and the sampling points; constructing a trajectory from the target node to the sampling points according to the trajectory search strategy of the fast expanding random tree, where the trajectory search strategy includes: if the target node is surrounded by obstacles, constructing the trajectory using a point-chasing method; otherwise, constructing the trajectory using a Durbins curve; determining new trajectory points based on the target node, the trajectory from the target node to the sampling points, and a preset driving distance; and adding the new trajectory points as new nodes to the fast expanding random tree. Determining a first planned trajectory from the vehicle's starting position to the target expansion position includes: determining the target expansion position based on the distances between the nodes of the fast-expanding random tree and each of the expansion positions; and determining the first planned trajectory from the starting position to the target expansion position based on the fast-expanding random tree. Determine a second planned trajectory from the target extension location to the parking location; Based on the first planned trajectory and the second planned trajectory, the planned trajectory from the starting position to the parking position is determined.

2. The parking trajectory planning method according to claim 1, characterized in that, The determination of multiple extended positions based on the relative positional relationship between the vehicle's starting position and parking position includes: Multiple driving trajectories are determined based on the relative positional relationship and the parking position, and the driving trajectories conform to the kinematic equations of the vehicle. The multiple driving trajectories are sampled to determine the multiple extended locations.

3. The parking trajectory planning method according to claim 2, characterized in that, The process of determining multiple driving trajectories based on the relative positional relationship and the parking position includes: The parking location type is determined based on the relative positional relationship; Multiple driving trajectories are determined based on the parking location type and the parking location.

4. The parking trajectory planning method according to claim 3, characterized in that, The process of determining multiple driving trajectories based on the parking location type and the parking location includes: If the parking location type is a perpendicular parking space, the first extended location is determined based on the first preset driving direction, the first preset driving distance, and the parking location; Multiple driving trajectories are generated based on multiple first preset attitude angles and the first extended position; The perpendicular parking space is either a rear-entry perpendicular parking space or a front-entry perpendicular parking space, and the first preset driving direction corresponding to the rear-entry perpendicular parking space is opposite to the first preset driving direction corresponding to the front-entry perpendicular parking space.

5. The parking trajectory planning method according to claim 3, characterized in that, The process of determining multiple driving trajectories based on the parking location type and the parking location includes: If the parking location type is a parallel parking space, the second extended location is determined based on the second preset driving direction, the second preset driving distance, and the parking location; The third extension position is determined based on the third preset driving direction, the third preset driving distance, and the second extension position; Multiple driving trajectories are generated based on multiple second preset attitude angles and the third extended position; The parallel parking space is either a first parallel parking space or a second parallel parking space. The parking direction of the vehicle corresponding to the first parallel parking space is opposite to that of the vehicle corresponding to the second parallel parking space. The third preset driving direction corresponding to the first parallel parking space is opposite to that of the third preset driving direction corresponding to the second parallel parking space.

6. The parking trajectory planning method according to claim 3, characterized in that, The process of determining multiple driving trajectories based on the parking location type and the parking location includes: If the parking location type is an angled parking space, the fourth extended location is determined based on the fourth preset driving direction, the fourth preset driving distance, and the parking location; Multiple driving trajectories are generated based on multiple third preset attitude angles and the fourth extended position; The inclined train position is either a rear-entry inclined train position or a front-entry inclined train position, and the fourth preset driving direction corresponding to the rear-entry inclined train position is opposite to the fourth preset driving direction corresponding to the front-entry inclined train position.

7. The parking trajectory planning method according to claim 1, characterized in that, The determination of multiple extended positions based on the relative positional relationship between the vehicle's starting position and parking position includes: Obtain obstacle information in the space where the vehicle is located; The plurality of extended positions are determined based on the obstacle information and the relative positional relationship.

8. The parking trajectory planning method according to claim 1, characterized in that, The determination of the first planned trajectory from the starting position to the target extended position includes: Based on the fast expanding random tree and parent node search algorithm, the first planned trajectory is backtracked from the target expansion position to the starting position.

9. The parking trajectory planning method according to claim 1, characterized in that, The second planned trajectory for determining the target extension location to the parking location includes: Determine the driving trajectory from the parking location to the target extended location; Based on the opposite direction of the driving trajectory, a second planned trajectory is determined, starting from the target expansion position and ending at the parking position.

10. The parking trajectory planning method according to claim 1, characterized in that, The trajectory search strategy of the fast expanding random tree conforms to the kinematic equations of the vehicle.

11. The trajectory planning method according to claim 10, characterized in that, The trajectory search strategy includes: If the trajectory endpoint includes attitude angle information, a trajectory from the trajectory start point to the trajectory endpoint is constructed based on the Durbins curve; and / or, If the trajectory endpoint does not include attitude angle information, the trajectory from the trajectory starting point to the trajectory endpoint is constructed based on the target attitude angle and the point chasing algorithm.

12. The parking trajectory planning method according to claim 11, characterized in that, The parking trajectory planning method also includes: The target attitude angle is determined based on the distance between the trajectory start point and the trajectory end point, the attitude angle of the trajectory start point, the wheelbase of the vehicle, and the vector angle from the trajectory start point to the trajectory end point.

13. The parking trajectory planning method according to claim 1, characterized in that, The step of sampling in the space where the vehicle is located to obtain sampling points includes: According to a preset sampling probability, sampling is performed in the space where the vehicle is located to obtain sampling points; the preset sampling probability is used to characterize the probability that the location corresponding to the sampling point belongs to the plurality of extended locations.

14. The parking trajectory planning method according to claim 1, characterized in that, The parking trajectory planning method also includes: Obtain obstacle information in the space where the vehicle is located; The step of adding the new trajectory point as a new node to the fast-expanding random tree includes: Based on the obstacle information, determine whether the trajectory from the target node to the new trajectory point contains obstacles; If the trajectory from the target node to the new trajectory point does not contain any obstacles, the new trajectory point is added as a new node to the fast-expanding random tree.

15. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When executed by a processor, the program instructions implement the steps of the method described in any one of claims 1 to 14.

16. A vehicle, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to perform the steps of the method according to any one of claims 1 to 14.

Images