Parking method, electronic equipment and storage medium

Abstract

The invention provides a parking method, electronic equipment and a storage medium, and the method comprises the steps: obtaining a current pose, parking reference information and a final parking pose of a vehicle, carrying out the geometric path planning according to the current pose, the parking reference information and the final parking pose, and obtaining a first parking path, under the condition that all the first parking paths do not meet the preset safety condition, path searching is carried out according to the parking reference information and a vehicle kinematics model, a second parking path from the current pose to the final parking pose is searched out, and in response to successful searching of the second parking path, the vehicle is controlled to be parked according to the second parking path. According to the method, the path planning complexity can be reduced by utilizing geometric path planning in a relatively simple parking scene, the parking time is reduced, the parking efficiency is improved, the parking safety is ensured by utilizing path searching in a relatively complex parking scene, and effective obstacle avoidance of moving obstacles is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of unmanned driving, in particular to a parking method, an electronic device and a storage medium. BACKGROUND

[0002] Autonomous parking planning is a key part of an unmanned driving system, and its essence is a path search problem, which provides a series of steering wheel turning angles and gear information for an intelligent driving vehicle from a starting pose to a target pose through certain rules.

[0003] At present, various methods have been studied for parking path planning, including curve fitting methods (such as spiral lines, spline interpolation, polynomials, etc.), geometric drawing methods, and search algorithms based on two-dimensional grid maps (such as Dijkstra, A*, etc.).

[0004] Among them, the parking path planned by the curve fitting method has problems such as difficulty in finding control points, complex path generation algorithm, and inability to effectively avoid obstacles for moving obstacles, and it is difficult to dynamically adjust the actual trajectory when there is an error between the actual trajectory and the planned trajectory; the geometric drawing method has a small applicable range, is not flexible, and has problems such as sudden change in curvature; the search algorithm based on the two-dimensional grid map takes much longer time to search the path than other heuristic search algorithms, and if the resolution of the two-dimensional model is set more finely, the real-time performance of the algorithm will face greater challenges due to the increase in search space. SUMMARY

[0005] In view of the above defects or deficiencies in the prior art, the present application aims to provide a parking method, an electronic device and a storage medium to solve the problems of complex path planning generation, small applicable range, sudden change in curvature, long time consumption and the like in the related art.

[0006] The embodiment of the present application provides a parking method, which comprises:

[0007] Obtaining a current pose of a vehicle, parking reference information and a final parking pose, and generating a first parking path of the vehicle based on the current pose, the parking reference information and the final parking pose;

[0008] In response to all first parking paths not meeting a preset safety condition, searching a second parking path from the current pose to the final parking pose based on the parking reference information and a kinematic model of the vehicle;

[0009] In response to the second parking path search being successful, controlling the vehicle to park based on the second parking path.

[0010] Optionally, the first parking path is a path obtained by connecting the current pose, poses of the sampling points, and the final parking pose, and the second parking path is a path obtained by searching from the current pose to the final parking pose.

[0011] Optionally, the parking reference information comprises map boundary and parking space information, and obtaining the final parking pose of the vehicle comprises:

[0012] determining a target position and a corresponding target direction in the map boundary based on a parking space vertex position and a parking space size in the parking space information;

[0013] determining the target position and the target direction as the final parking pose.

[0014] Optionally, the parking reference information comprises map boundary and parking space information, and generating the first parking path of the vehicle based on the current pose, the parking reference information, and the final parking pose comprises:

[0015] determining a pose adjustment area of the vehicle in the map boundary based on the current pose and the final parking pose, and determining a plurality of sampling points in the pose adjustment area;

[0016] for each of the sampling points, determining a pose of the sampling point according to a positional relationship between the sampling point and a parking space vertex in the parking space information;

[0017] generating an advancing path in the first parking path according to the current pose and the pose of the sampling point, and generating a retreating path in the first parking path according to the pose of the sampling point and the final parking pose.

[0018] Optionally, the parking reference information comprises obstacle information, and after the first parking path of the vehicle is generated, the method further comprises:

[0019] determining whether the first parking path satisfies the preset safety condition;

[0020] wherein determining whether the first parking path satisfies the preset safety condition comprises:

[0021] performing collision detection on the first parking path based on the obstacle information;

[0022] in response to detecting that the first parking path has a collision risk, determining that the first parking path does not satisfy the preset safety condition.

[0023] Optionally, after performing collision detection on the first parking path based on the obstacle information, the method further comprises:

[0024] In response to detecting that the first parking path has no collision risk and the curvature of the first parking path is greater than the preset curvature threshold, it is determined that the first parking path meets the preset safety condition.

[0025] Optionally, after determining whether the first parking path meets the preset safety condition, the method further comprises:

[0026] If the first parking path meets the preset safety condition, the path length and the path cumulative curvature of the first parking path are determined.

[0027] According to the path length and the path cumulative curvature, a path operation score of the first parking path is determined.

[0028] In response to the existence of the first parking path meeting the preset safety condition, a target parking path is determined according to the path operation score of the first parking path meeting the preset safety condition, and the vehicle is controlled to park according to the target parking path.

[0029] Optionally, the parking reference information comprises map boundary and obstacle information, and a second parking path from the current pose to the final parking pose is searched based on the parking reference information and a vehicle kinematics model, comprising:

[0030] An occupancy grid is constructed based on the map boundary and the obstacle information, and a passable area is marked in the occupancy grid.

[0031] The current pose is taken as a node to be searched, and a next optimal node of the node to be searched is searched in the passable area based on the vehicle kinematics model, wherein the vehicle kinematics model describes the relationship between the pose of the vehicle and the control input, and the vehicle kinematics model is used to determine a next node according to the node to be searched and the control input.

[0032] The next optimal node is taken as a new node to be searched, and the step of searching the next optimal node of the node to be searched in the passable area based on the vehicle kinematics model is returned until the final parking pose is reached, so as to obtain the second parking path from the current pose to the final parking pose.

[0033] Optionally, before controlling the vehicle to park based on the second parking path, the method further comprises:

[0034] A preset path length and a preset curvature change rate are obtained.

[0035] The second parking path is adjusted to update the second parking path, with a path length of the second parking path not exceeding the preset path length and a curvature change rate of the second parking path not exceeding the preset curvature change rate as a target.

[0036] The electronic device provided by the embodiment of the present application comprises:

[0037] a processor and a memory;

[0038] The processor is configured to execute the steps of the parking method provided by any of the embodiments of the present application by invoking programs or instructions stored in the memory.

[0039] The present application also provides a computer readable storage medium storing programs or instructions, which make a computer execute the steps of the parking method provided by any of the embodiments of the present application.

[0040] In summary, the present application provides a parking method, which obtains a current pose of a vehicle, parking reference information and a final parking pose, performs geometric path planning according to the current pose, the parking reference information and the final parking pose to obtain a first parking path, performs path searching according to the parking reference information and a kinematic model of the vehicle to search for a second parking path from the current pose to the final parking pose in a case where all the first parking paths do not meet preset safety conditions, and controls the vehicle to park according to the second parking path in response to the second parking path searching being successful, so as to realize automatic parking of the vehicle. The method first obtains the first parking path through geometric path planning, and then obtains the second parking path through searching in a case where all the first parking paths do not meet the preset safety conditions. The geometric path planning can be used to reduce path planning complexity, ensure parking stability, reduce parking time consumption and improve parking efficiency in a relatively simple parking scenario. In addition, the path searching can be used to ensure parking safety, avoid the curvature mutation path of the geometric path planning and realize effective obstacle avoidance of a moving obstacle in a relatively complex parking scenario, so as to expand the application range of path planning. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed in the description of the embodiments or the prior art will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0042] Figure 1 is a flowchart of a parking method provided by an embodiment of the present application;

[0043] Figure 2 is a schematic diagram of geometric path planning provided by an embodiment of the present application;

[0044] Figure 3 is a schematic diagram of smoothing of a second parking path provided by an embodiment of the present application;

[0045] Figure 4 is a flowchart of parking path planning provided by an embodiment of the present application;

[0046] Figure 5 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0047] The present application will be further described below in conjunction with the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related application, and not to limit the application. In addition, it should be noted that only the parts related to the application are shown in the drawings for ease of description.

[0048] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0049] As mentioned in the background, in view of the problems in the prior art, the present application proposes a parking method. Figure 1 is a flowchart of a parking method provided by an embodiment of the present application. Referring to Figure 1 , the parking method specifically includes:

[0050] S110, obtaining a current pose of the vehicle, parking reference information, and a final parking pose, and generating a first parking path of the vehicle based on the current pose, the parking reference information, and the final parking pose.

[0051] The current pose can be the pose of the vehicle at the current time, including the position of the vehicle and the direction (such as the yaw angle) of the vehicle. The parking reference information can be the reference information required for the vehicle to plan a parking path, including map boundaries, parking space information, obstacle information, etc.

[0052] In the embodiments of the present application, the current pose and the parking reference information of the vehicle can be obtained through the vehicle sensors. Specifically, the current pose, the parking reference information, and the final parking pose of the vehicle can be obtained when it is detected that the vehicle meets the parking path planning condition; wherein the parking path planning condition can be that the vehicle reaches the navigation destination or the user actively triggers the parking function (voice trigger, trigger through a button, or trigger in the car screen).

[0053] The final parking pose can be a pose at which the vehicle completes parking into the parking space, i.e., a target pose, including a position of the vehicle and a direction of the vehicle. The final parking pose can be calculated in real time according to the size of the parking space, the vertex of the parking space, and the like.

[0054] In a specific embodiment, the parking reference information includes a map boundary and parking space information, and the final parking pose of the vehicle is obtained, including:

[0055] According to the vertex position of the parking space and the size of the parking space in the parking space information, a target position and a corresponding target direction are determined in the map boundary; and the target position and the target direction are determined as the final parking pose.

[0056] The vertex position of the parking space can be a position of a vertex of the parking space in the parking map, and the map boundary can be a boundary of the parking map. Specifically, the final parking position, i.e., the target position, can be planned within the map boundary according to the vertex position of the parking space and the size of the parking space, and the corresponding target direction is determined. The target direction can be a travel direction of the vehicle at the target position, i.e., a yaw angle of the vehicle.

[0057] For example, the vertex position of the parking space, the size of the parking space, the map boundary, the size of the vehicle, and the type of the parking space can be used as prior information, a coordinate system is established with the geometric center or the vertex of the parking space as the origin, and then the target position within the map boundary is calculated according to the type of the parking space, and the target direction is determined according to the final parking position and the orientation of the parking space, and then the target position and the target direction are determined as the final parking pose.

[0058] After the final parking pose is determined, a safety buffer can also be added on the basis of the final parking pose, for example, whether the distance between the target position and the edge of the parking space is greater than a preset safety margin is determined, and if not, the target position or the target direction is adjusted to further improve the parking safety.

[0059] In the embodiments of the present application, after the current pose, the final parking pose, and the parking reference information are obtained, the current pose, the final parking pose, and the parking reference information are used to perform geometric path planning to obtain a first parking path of the vehicle. The geometric path planning can be to generate a global path through geometric construction, and specifically, the first parking path can be formed by connecting the starting point (the current position of the vehicle), the target point (the final parking position of the vehicle), and the sampling point.

[0060] In a specific embodiment, the first parking path is obtained by connecting the current pose, the pose of the sampling point, and the final parking pose, and the second parking path is obtained by searching from the current pose to the final parking pose.

[0061] The first parking path can be obtained by connecting the current pose and the sampling points, and connecting the sampling points and the final parking pose. In the embodiments of the present application, the sampling points can be position points determined within the map boundary for adjusting the vehicle pose in the parking process.

[0062] The second parking path can be obtained by searching from the current pose to the final parking pose. Specifically, the next pose of the current pose can be searched from the current pose, and then the next pose is continuously searched from the next pose, and the process is repeated until the final parking pose is searched. The current pose, all searched poses, and the final parking pose are connected to obtain the second parking path.

[0063] Since the second parking path is obtained by searching each pose from the current pose, and the adjacent poses are connected in sequence, the adjacent poses searched in this process are continuously changed, that is, the turning radius between the adjacent poses is continuously changed, and a curve with continuously changing turning radius can be obtained by connecting each pose in sequence. Therefore, the second parking path can be understood as a continuous parking path, which is smooth in geometry and can enable the vehicle to smoothly track the parking path by front wheel steering, ensuring the smoothness and safety of parking.

[0064] In the embodiments of the present application, all first parking paths are generated first, and the second parking path is generated in the case that all first parking paths do not meet the preset safety condition. The purpose is to consider that the first parking path can be obtained by connecting the current pose, the pose of the sampling points, and the final parking pose, without the need to search the pose combined with the vehicle kinematics model, which can ensure the efficiency of path generation. Therefore, the first parking path can be generated first to reduce the time consumption of path generation, thereby improving the parking efficiency. In the case that all first parking paths do not meet the preset safety condition, the second parking path is further generated by searching, which realizes accurate planning of the parking path and solves the problem of unstable geometric path planning, while ensuring the parking stability while considering the parking efficiency.

[0065] In some embodiments, the parking reference information includes the map boundary and the parking space information, and the first parking path of the vehicle is generated based on the current pose, the parking reference information, and the final parking pose, including the following steps:

[0066] Step 11, determining a pose adjustment region of the vehicle in the map boundary based on the current pose and the final parking pose, and determining a plurality of sampling points in the pose adjustment region;

[0067] Step 12, for each sampling point, determining the pose of the sampling point according to the positional relationship between the sampling point and the vertex of the parking space in the parking space information;

[0068] Step 13, generating a forward path in the first parking path according to the current pose and the poses of the sampling points, and generating a backward path in the first parking path according to the poses of the sampling points and the final parking pose.

[0069] In step 11, the pose adjustment area of the vehicle in the map boundary can be determined according to the current pose and the final parking pose, in combination with the type of the parking space. The pose adjustment area can be understood as an area in which the vehicle adjusts the pose before parking in the parking space. For example, the pose adjustment area can be an area for adjusting the direction of the vehicle to be close to the direction of the vehicle in the final parking pose. The area covered by the vehicle driving straight or turning along the direction of the vehicle in the final parking pose can be determined as the pose adjustment area.

[0070] For example, Figure 2 is a schematic diagram of a geometric path planning provided by an embodiment of the present application, as Figure 2 shown, the line segment AB, the line segment BC, the line segment CD, the line segment DE, the line segment EF, the line segment FG, the line segment GH, and the line segment HA constitute the map boundary. Within the map boundary, the pose adjustment area can be determined to be located in front of and to the upper right of the end point from the start point of the vehicle. For example, the rectangular area constituted by the line segment DE and the line segment EF can be regarded as a part of the pose adjustment area.

[0071] After the pose adjustment area is determined, further sampling can be performed in the pose adjustment area to obtain a plurality of sampling points. For example, the sampling interval distance corresponding to the length and the width of the pose adjustment area can be determined according to the length and the width of the pose adjustment area, and then a plurality of coordinate points, i.e., sampling points, can be obtained by sampling in the pose adjustment area according to the sampling interval distance corresponding to the length and the width. Figure 2 For example, as

[0072] Further, in step 12, for each sampling point, the positional relationship between the sampling point and the parking space vertex can be determined. The positional relationship can be determined by the direction of the sampling point relative to the parking space vertex. For example, as Figure 2 shown, the sampling point 1, the sampling point 2, and the sampling point 3 are all located to the upper right of the parking space vertex A.

[0073] After the positional relationship between the sampling point and the parking space vertex is determined, the sampling angle range can be obtained according to the positional relationship. For example, a rectangular coordinate system can be constructed with the parking space vertex as the origin, and then the corresponding right angle range in the rectangular coordinate system can be determined as the angle sampling range according to the direction of the sampling point relative to the parking space vertex. After the sampling angle range is obtained, sampling can be performed according to the interval angle to obtain the pose of the sampling point. For example, Figure 2For example, the sampling point is located at the right upper corner of the parking space vertex A, and thus the sampling angle range can be φ.

[0074] Further, in step 13, each sampling point can be traversed to generate a first parking path corresponding to each sampling point in sequence. Specifically, all sampling points can be stored in a list in sequence, and then the first sampling point in the list is taken as a current sampling point to generate a first parking path corresponding to the current sampling point, and then the next sampling point of the current sampling point in the list is taken as a new current sampling point to generate a first parking path corresponding to the current sampling point, and the process is repeated until all first parking paths corresponding to the sampling points in the list are obtained.

[0075] Specifically, for each sampling point in the list, the following operations can be performed: generating a forward path in the first parking path according to the current pose and the pose of the sampling point, for example, connecting the current pose and the pose of the sampling point by a quintic polynomial to obtain the forward path, and generating a backward path in the first parking path according to the pose of the sampling point and the final parking pose, for example, connecting the pose of the sampling point and the final parking pose by a quintic polynomial to obtain the backward path.

[0076] By repeatedly performing the above operations, a first parking path corresponding to each sampling point can be obtained. As shown in Figure 2 The first parking path includes a forward path and a backward path, the forward path corresponding to the sampling point 1 is a route from the starting point to the sampling point 1, and the backward path corresponding to the sampling point 1 is a route from the sampling point 1 to the ending point; the forward path corresponding to the sampling point 2 is a route from the starting point to the sampling point 2, and the backward path corresponding to the sampling point 2 is a route from the sampling point 2 to the ending point; the forward path corresponding to the sampling point 3 is a route from the starting point to the sampling point 3, and the backward path corresponding to the sampling point 3 is a route from the sampling point 3 to the ending point.

[0077] By the above steps 11-13, the position adjustment region in the vehicle parking process can be determined first, and then sampling is performed in the region, and a plurality of first parking paths are obtained through different sampling points, which can ensure the comprehensiveness of the geometric path planning, and the generation process is simple, which can improve the efficiency of path planning.

[0078] S120, in response to all first parking paths not satisfying the preset safety condition, searching a second parking path from the current pose to the final parking pose based on the parking reference information and the vehicle kinematics model.

[0079] In the embodiments of the present application, after obtaining each first parking path of the vehicle, all the first parking paths can be traversed to determine whether each first parking path satisfies a preset safety condition. The preset safety condition can be a condition preset to ensure the safety of the vehicle in parking, for example, the preset safety condition can include that the first parking path does not have a collision risk.

[0080] In some embodiments, the parking reference information includes obstacle information, and after generating the first parking path of the vehicle, the method further includes determining whether the first parking path satisfies a preset safety condition.

[0081] The determination of whether the first parking path satisfies the preset safety condition includes: performing collision detection on the first parking path based on the obstacle information; and in response to detecting that the first parking path has a collision risk, determining that the first parking path does not satisfy the preset safety condition.

[0082] Specifically, after the geometric path planning, each first parking path can be sequentially subjected to collision detection according to the obstacle information, wherein the obstacle information can include an obstacle type and an obstacle position.

[0083] For example, according to the obstacle type and the obstacle position, the distance between the obstacle and the first parking path can be determined, and then it is determined whether there is a collision risk according to the distance. If there is, it is determined that the first parking path satisfies the preset safety condition, and the first parking path is stored. If there is not, it is determined that the first parking path does not satisfy the preset safety condition, and the next first parking path is determined whether it satisfies the preset safety condition.

[0084] In the above embodiments, whether the first parking path satisfies the preset safety condition can be determined according to whether the first parking path has a collision risk, thereby avoiding collision between the vehicle and the obstacle during parking.

[0085] In addition to whether the first parking path has a collision risk, if the curvature of the first parking path is too large when the vehicle travels along the first parking path, there will also be a risk of driving safety. Therefore, the preset safety condition can include that the first parking path does not have a collision risk and the curvature of the first parking path is less than a preset curvature threshold.

[0086] In an example, after performing collision detection on the first parking path based on the obstacle information, the method further includes:

[0087] In response to detecting that the first parking path does not have a collision risk and the curvature of the first parking path is greater than the preset curvature threshold, it is determined that the first parking path satisfies the preset safety condition.

[0088] The preset curvature threshold can be a preset maximum curvature limit value, which can be determined according to the turning ability of the vehicle and the maximum curvature that guarantees safe driving of the vehicle.

[0089] Specifically, if it is detected that the first parking path does not have a collision risk, it can be further determined whether the curvature of the first parking path is greater than the preset curvature threshold. If yes, it can be determined that the first parking path does not meet the preset safety condition. If no, it can be determined that the first parking path meets the preset safety condition.

[0090] In the above example, in addition to determining whether the first parking path has a collision risk, it can also be determined whether the curvature of the first parking path is greater than the preset curvature threshold, so that the first parking path that does not have a collision risk and whose curvature does not exceed the preset curvature threshold can be screened out, further ensuring the safety of vehicle parking.

[0091] In the embodiments of the present application, after it is determined whether each first parking path meets the preset safety condition, the first parking path that meets the preset safety condition can be stored. Further, if the number of stored first parking paths is zero, it indicates that all first parking paths do not meet the preset safety condition, at which time path searching can be further performed, and the searching method can be hybrid A star.

[0092] In the embodiments of the present application, if all first parking paths do not meet the preset safety condition, the purpose of using hybrid A star to further implement path searching is that hybrid A star generates successor nodes by integrating a vehicle kinematics model, which can simulate real forward, backward and turning actions of the vehicle, so that the generated second parking path is continuous and has smooth curvature, i.e., the smoothness of the path is ensured, so that the path meets the turning radius limit of the vehicle, which is convenient for subsequent vehicle to directly track and execute, improves parking safety, and is suitable for various complex parking scenarios, such as narrow parking spaces, complex turning areas, and unstructured environments full of obstacles.

[0093] Specifically, the passable area of the vehicle can be determined within the map boundary according to the obstacle information, and then a second parking path from the current pose to the final parking pose can be searched out in the passable area.

[0094] In some embodiments, the parking reference information includes a map boundary and obstacle information, and a second parking path from a current pose to a final parking pose is searched out based on the parking reference information and a vehicle kinematics model, including the following steps:

[0095] Step 21, constructing an occupancy grid based on the map boundary and the obstacle information, and marking a passable area in the occupancy grid;

[0096] Step 22, taking the current pose as a to-be-searched node, searching for a next optimal node of the to-be-searched node in the passable area based on a vehicle kinematics model, wherein the vehicle kinematics model describes a relationship between a pose of the vehicle and a control input, and the vehicle kinematics model is used to determine the next node according to the to-be-searched node and the control input;

[0097] Step 23, taking the next optimal node as a new to-be-searched node, returning to the step of searching for a next optimal node of the to-be-searched node in the passable area based on the vehicle kinematics model until a final parking pose is reached, thereby obtaining a second parking path from the current pose to the final parking pose.

[0098] In step 21, an occupancy grid can be constructed within the map boundary according to the obstacle positions and the obstacle types in the obstacle information, and the obstacle positions and the adjacent areas are marked as impassable areas, so that other areas within the map boundary are marked as passable areas.

[0099] The size of the adjacent area of the obstacle that needs to be marked as an impassable area can be determined according to the obstacle type, and specifically, a corresponding reserved safety distance can be determined according to the obstacle type, for example, if the obstacle type is a dynamic obstacle, the larger the reserved safety distance, and then the adjacent area that needs to be marked as an impassable area can be determined according to the reserved safety distance.

[0100] Further, in step 22, the current pose can be taken as a to-be-searched node, and a next optimal node of the to-be-searched node can be searched in the passable area according to the vehicle kinematics model.

[0101] For example, the vehicle kinematics model can determine each next node of the to-be-searched node according to the to-be-searched node and each control input, and then determine the cost of each next node, and determine the node with the minimum cost as the optimal node. The cost can include path length cost, steering angle cost, reverse penalty cost, direction switching cost, and environment cost.

[0102] The path length cost can be positively correlated with the path length, the steering angle cost can be positively correlated with the change amount of the steering angle, the reverse penalty cost can be positively correlated with the number of times of reversing of the vehicle, the direction switching cost can be positively correlated with the number of times of direction switching, and the environment cost can include collision risk cost and terrain cost, the collision risk cost is negatively correlated with the distance of the path relative to the obstacle, and the terrain cost can be increased when the ground is rough or the ground is sloped.

[0103] After the next optimal node is searched, the next optimal node can be re-set as a new node to be searched, and the step of searching in the passable area through the vehicle kinematic model is executed again to obtain a next optimal node of the new node to be searched, and the process is repeated until the final parking pose is reached, and a second parking path from the current pose to the final parking pose is obtained.

[0104] Through steps 21-23, the path search combined with kinematic constraints can be realized, which is suitable for complex parking scenarios such as narrow space and dynamic obstacles, and can ensure the feasibility of the generated path while taking into account the path smoothness.

[0105] In addition, steps 21-23 search nodes through the vehicle kinematic model, which can simulate the real forward, backward and turning actions of the vehicle, so that the adjacent nodes are continuous in the state space, i.e., there is no direction mutation between adjacent nodes. Therefore, the second parking path generated by connecting all optimal nodes in sequence is curvature-continuous, i.e., a curvature-continuous second parking path is obtained. This way can ensure the smoothness of the second parking path, so that the second parking path meets the turning radius limit of the vehicle, and the vehicle can directly track and execute the second parking path, ensuring the smoothness and safety of parking.

[0106] S130, in response to the success of the second parking path search, controlling the vehicle to park based on the second parking path.

[0107] Specifically, if the second parking path search is successful, the second parking path can be issued to other control domains to control the vehicle to park based on the second parking path.

[0108] Considering that there may be no feasible path in the environment due to the parking space being surrounded by obstacles or being unable to pass through narrow spaces under kinematic constraints, and thus the second parking path search fails, in response to the failure of the second parking path search, corresponding prompt information can be generated and displayed through the vehicle interface or voice device to remind the user that the parking path planning fails. The prompt information can include failure root cause information of the parking path search, so that the user can determine whether to change the parking space through the failure root cause information.

[0109] In the embodiments of the present application, if the second parking path is searched, in order to further ensure the executability, comfort and parking efficiency of the second parking path, the second parking path can also be smoothed.

[0110] In some embodiments, before controlling the vehicle to park based on the second parking path, the following steps are further included:

[0111] Step 31, obtaining a preset path length and a preset curvature change rate;

[0112] Step 32, adjust the second parking path to update the second parking path, with the path length of the second parking path not exceeding a preset path length, and the curvature change rate of the second parking path not exceeding a preset curvature change rate as a target.

[0113] In step 31, the preset path length and the preset curvature change rate can be obtained first. The preset path length can be a preset parking driving path length limit value, and the preset curvature change rate can be a preset curvature change limit value of the parking driving path.

[0114] Specifically, in step 32, the path length of the second parking path not exceeding the preset path length and the curvature change rate of the second parking path not exceeding the preset curvature change rate can be set as a smoothing target. The second parking path is adjusted through the smoothing target, so that the adjusted second parking path meets the smoothing target, and the purpose of updating the second parking path is achieved.

[0115] For example, the position of the path point in the second parking path can be adjusted to reduce the curvature change, or a smoothing curve can be used to fit the second parking path in a spline interpolation manner.

[0116] Through the above embodiments, the second parking path can be smoothed after being searched, so as to solve the problems of sawtooth jitter, curvature mutation, and redundant node increase in the path length of the searched second parking path, and ensure the executability, comfort, and efficiency of the second parking path.

[0117] Figure 3 is a smoothing diagram of a second parking path provided by an embodiment of the present application. As shown in Figure 3 Compared with the searched second parking path, the curvature of the smoothed second parking path is smaller, and the path length is shorter.

[0118] In the embodiments of the present application, it should be noted that after it is judged whether each first parking path meets the preset safety condition, if the first parking path is stored, it means that there is a first parking path meeting the preset safety condition, and one of the first parking paths can be selected to control the vehicle to park. The final first parking path can be selected according to the score of the first parking path, and the score can be evaluated by the path length and the path cumulative curvature when it is judged that the first parking path meets the preset safety condition.

[0119] In an example, after it is judged whether the first parking path meets the preset safety condition, the following step is further included:

[0120] Step 41, if the first parking path meets the preset safety condition, the path length and the path cumulative curvature of the first parking path are determined.

[0121] Step 42: Determine the path operation score for the first parking path based on the path length and cumulative path curvature;

[0122] Step 43: In response to the existence of a first parking path that meets the preset safety conditions, determine the target parking path based on the path operation score of the first parking path that meets the preset safety conditions, and control the vehicle to park according to the target parking path.

[0123] In step 41, after determining whether the first parking path meets the preset safety conditions, if the first parking path meets the preset safety conditions, the path length and cumulative curvature of the first parking path can be determined. The cumulative curvature can be the sum of the curvatures of all points in the first parking path.

[0124] Furthermore, in step 42, the parking efficiency and parking feasibility of the first parking path can be evaluated using the path length and cumulative path curvature to obtain a path operation score. For example, the path operation score can be obtained by weighting the path length and cumulative path curvature using preset length weights and preset curvature weights.

[0125] After obtaining the path operation scores of all first parking paths that meet the preset safety conditions, a target parking path can be selected from each first parking path based on the path operation scores of the first parking paths that meet the preset safety conditions, and then the vehicle can be parked according to the target parking path.

[0126] The above example demonstrates that, provided the first parking path meets preset safety conditions, the first parking path can be selected based on its length and cumulative curvature, ensuring both efficiency and feasibility of vehicle parking.

[0127] Figure 4 This is a flowchart of a parking path planning method provided in an embodiment of this application, such as... Figure 4 As shown, firstly, map boundaries, parking space information, obstacle information, and vehicle parameters can be obtained to determine the parking start point and parking end point, that is, to determine the current pose and the final parking pose. Then, a first parking path is generated through a geometric algorithm, and it is determined whether the first parking path meets the preset safety conditions. If there is a first parking path that meets the preset safety conditions, the parking path planning can be determined to be successful, and the process ends.

[0128] If no first parking path that meets the preset safety conditions exists, a second parking path from the parking start point to the parking end point is searched using the hybrid A-star algorithm. The success of the hybrid search is then determined. If it fails, the parking path planning is determined to have failed, and the process ends. If the hybrid search is successful, the second parking path is input into the downstream smoothing algorithm.

[0129] After the second parking path is input to the smoothing algorithm downstream, it can be further determined whether the smoothing processing is successful. If successful, it is determined that the parking path planning is successful, and the process is ended. If unsuccessful, the second parking path can be directly output, and the process is ended.

[0130] The parking method provided by the embodiments of the present application acquires the current pose of the vehicle, the parking reference information, and the final parking pose, performs geometric path planning according to the current pose, the parking reference information, and the final parking pose, obtains the first parking path, in the case that all the first parking paths do not meet the preset safety condition, performs path searching according to the parking reference information and the kinematic model of the vehicle, searches out the second parking path from the current pose to the final parking pose, and controls the vehicle to park according to the second parking path in response to the successful searching of the second parking path, so as to realize automatic parking of the vehicle. The method first obtains the first parking path through geometric path planning, and then obtains the second parking path through searching in the case that all the first parking paths do not meet the preset safety condition. The geometric path planning can be used to reduce the path planning complexity, ensure parking stability, reduce parking time, and improve parking efficiency in a relatively simple parking scene. In addition, the path searching can be used to ensure parking safety in a relatively complex parking scene, avoid the use of the curvature mutation path of the geometric path planning, and effectively avoid obstacles of moving obstacles, so as to expand the application range of path planning.

[0131] The embodiments of the present application further provide a parking device, which comprises a geometric planning module, a hybrid searching module, and a parking control module, wherein:

[0132] The geometric planning module is configured to acquire the current pose of the vehicle, the parking reference information, and the final parking pose, and generate the first parking path of the vehicle based on the current pose, the parking reference information, and the final parking pose.

[0133] The hybrid searching module is configured to search out the second parking path from the current pose to the final parking pose based on the parking reference information and the kinematic model of the vehicle in response to the fact that all the first parking paths do not meet the preset safety condition.

[0134] The parking control module is configured to control the vehicle to park based on the second parking path in response to the successful searching of the second parking path.

[0135] Optionally, the parking reference information comprises map boundaries and parking space information, and the geometric planning module is specifically configured to:

[0136] determine a target position and a corresponding target direction in the map boundary based on the position of the parking space vertex and the size of the parking space in the parking space information; and determine the target position and the target direction as a final parking pose.

[0137] In the above embodiments, the parking reference information includes map boundary and parking space information, and the geometry planning module is specifically configured to:

[0138] determine a pose adjustment area of the vehicle in the map boundary based on the current pose and the final parking pose, and determine a plurality of sampling points in the pose adjustment area;

[0139] for each of the sampling points, determine a pose of the sampling point according to the positional relationship between the sampling point and the position of the parking space vertex in the parking space information;

[0140] generate a forward path in the first parking path according to the current pose and the pose of the sampling point, and generate a backward path in the first parking path according to the pose of the sampling point and the final parking pose.

[0141] In the above embodiments, the parking reference information includes obstacle information, and the geometry planning module is further configured to:

[0142] after generating the first parking path of the vehicle, determine whether the first parking path meets the preset safety condition;

[0143] The geometry planning module is further configured to:

[0144] perform collision detection on the first parking path based on the obstacle information; and in response to detecting that the first parking path has a collision risk, determine that the first parking path does not meet the preset safety condition.

[0145] In the above embodiments, the geometry planning module is further configured to:

[0146] in response to detecting that the first parking path has no collision risk, and the curvature of the first parking path is greater than the preset curvature threshold, determine that the first parking path meets the preset safety condition.

[0147] In the above embodiments, the geometry planning module is further configured to:

[0148] if the first parking path meets the preset safety condition, determine a path length and a path cumulative curvature of the first parking path;

[0149] determine a path operation score of the first parking path according to the path length and the path cumulative curvature.

[0150] In response to the existence of the first parking path satisfying the preset safety condition, a target parking path is determined according to a path operation score of the first parking path satisfying the preset safety condition, and the vehicle is controlled to park according to the target parking path.

[0151] On the basis of the above-mentioned embodiments, optionally, the parking reference information comprises map boundary and obstacle information, and the hybrid search module is specifically configured to:

[0152] construct an occupancy grid based on the map boundary and the obstacle information, and mark a passable area in the occupancy grid;

[0153] take the current pose as a to-be-searched node, search a next optimal node of the to-be-searched node in the passable area based on a vehicle kinematics model, wherein the vehicle kinematics model describes a relationship between a pose of a vehicle and a control input, and the vehicle kinematics model is used to determine a next node according to the to-be-searched node and a control input;

[0154] take the next optimal node as a new to-be-searched node, return to the step of searching the next optimal node of the to-be-searched node in the passable area based on the vehicle kinematics model, until the final parking pose is reached, and obtain a second parking path from the current pose to the final parking pose.

[0155] On the basis of the above-mentioned embodiments, optionally, the hybrid search module is further configured to:

[0156] obtain a preset path length and a preset curvature change rate;

[0157] adjust the second parking path to update the second parking path, with the path length of the second parking path not exceeding the preset path length and the curvature change rate of the second parking path not exceeding the preset curvature change rate as a target.

[0158] The parking device provided by the embodiments of the present application can execute the steps in the parking method provided by the method embodiments of the present application, and has the execution steps and beneficial effects which will not be repeated here.

[0159] Figure 5 is a structural schematic diagram of an electronic device provided by the embodiments of the present application. As shown in Figure 5 The electronic device 400 comprises one or more processors 401 and a memory 402.

[0160] The processor 401 can be a central processing unit (CPU) or other form of processing unit having data processing and / or instruction executing capabilities, and can control other components in the electronic device 400 to perform desired functions.

[0161] The memory 402 can include one or more computer program products that can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory, for example, can include random access memory (RAM), cache memory, and / or the like. The non-volatile memory, for example, can include read-only memory (ROM), hard disk, flash memory, and / or the like. One or more computer program instructions can be stored on the computer-readable storage media, and the processor 401 can run the program instructions to implement the parking method of any embodiment of the present application described above and / or other desired functions. Various contents such as initial extrinsic parameters, threshold values, and the like can also be stored in the computer-readable storage media.

[0162] In one example, the electronic device 400 can further include an input device 403 and an output device 404, which are interconnected through a bus system and / or other forms of connection mechanisms (not shown). The input device 403 can include, for example, a keyboard, a mouse, and / or the like. The output device 404 can output various information to the outside, including pre-warning prompt information, braking force, and the like. The output device 404 can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and / or the like.

[0163] Of course, in order to simplify, Figure 5 Only some of the components in the electronic device 400 related to the present application are shown in FIG. 4, and components such as buses, input / output interfaces, and the like are omitted. In addition, the electronic device 400 can include any other appropriate components according to specific application cases.

[0164] In addition to the above method and device, an embodiment of the present application can also be a computer program product including computer program instructions, which, when run by a processor, cause the processor to perform the steps of the parking method provided by any embodiment of the present application.

[0165] The computer program product can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++, or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device or entirely on the remote computing device or server. The embodiments of the present application are not limited by the

[0166] Furthermore, embodiments of the present application can also be a computer readable storage medium, having stored thereon computer program instructions which, when executed by a processor, cause the processor to carry out the steps described in any embodiment of the method of parking according to the present application.

[0167] The computer readable storage medium can be a combination of one or more computer readable media. The computer readable media can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium can include, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium include the following: an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0168] It should be noted that the terms "comprises", "comprising", "includes", "including", or any other variants thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises a", "comprising", "includes", "including", or "has" does not, without more constraints, preclude the existence of additional identical elements other than the listed elements.

[0169] It should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", and the like, indicate relative positions or orientations based on the positions or orientations shown in the drawings, and are used only to facilitate the description of the present application and simplify the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. Unless otherwise expressly specified and limited, the terms "mounting", "connecting", "connecting" and the like should be broadly understood, for example, can be fixedly connected, can also be detachably connected, or integrally connected; can be mechanically connected, can also be electrically connected; can be directly connected, can also be indirectly connected through an intermediate medium, can be internal communication of two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances.

[0170] The principles and implementation modes of the present application are described herein by applying specific examples, and the above example descriptions are only used to help understand the method of the present application and its core idea. The above description is only the preferred embodiment of the present application. It should be pointed out that due to the limitation of language expression, there are infinite specific structures objectively, and for those skilled in the art, some improvements, refinements or changes can be made without departing from the principles of the present application, and the above technical features can also be combined in an appropriate manner; these improvements, refinements, changes or combinations, or the direct application of the inventive concept and technical solution to other occasions without improvement, should be regarded as the protection scope of the present application.

Claims

1. A parking method, characterized in that, include: The vehicle's current pose, parking reference information, and final parking pose are obtained, and a first parking path for the vehicle is generated based on the current pose, the parking reference information, and the final parking pose. In response to the fact that all first parking paths do not meet the preset safety conditions, a second parking path from the current pose to the final parking pose is searched based on the parking reference information and the vehicle kinematics model. In response to the successful search of the second parking path, the vehicle is controlled to park based on the second parking path.

2. The method according to claim 1, characterized in that, The first parking path is the path obtained by connecting the current pose, the pose of the sampling point, and the final parking pose. The second parking path is the path obtained by searching from the current pose to the final parking pose.

3. The method according to claim 1, characterized in that, The parking reference information includes map boundaries and parking space information, and the final parking position of the vehicle is obtained, including: Based on the parking space vertex position and parking space size in the parking space information, the target position and corresponding target direction are determined in the map boundary; The target position and the target direction are determined as the final parking position.

4. The method according to claim 1, characterized in that, The parking reference information includes map boundaries and parking space information. Based on the current pose, the parking reference information, and the final parking pose, a first parking path for the vehicle is generated, including: Based on the current pose and the final parking pose, the pose adjustment area of ​​the vehicle is determined in the map boundary, and multiple sampling points are determined in the pose adjustment area. For each of the sampling points, the pose of the sampling point is determined based on the positional relationship between the sampling point and the vertex of the parking space in the parking space information; A forward path in the first parking path is generated based on the current pose and the pose of the sampling point, and a backward path in the first parking path is generated based on the pose of the sampling point and the final parking pose.

5. The method according to claim 1, characterized in that, The parking reference information includes obstacle information, and after generating the first parking path for the vehicle, it also includes: Determine whether the first parking path meets the preset safety conditions; The determination of whether the first parking path meets the preset safety conditions includes: Collision detection is performed on the first parking path based on the obstacle information; In response to the detection of a collision risk in the first parking path, it is determined that the first parking path does not meet the preset safety conditions.

6. The method according to claim 5, characterized in that, After performing collision detection on the first parking path based on the obstacle information, the method further includes: In response to the detection that there is no collision risk in the first parking path and that the curvature of the first parking path is greater than the preset curvature threshold, it is determined that the first parking path meets the preset safety conditions.

7. The method according to claim 5, characterized in that, After determining whether the first parking path meets the preset safety conditions, the method further includes: If the first parking path meets the preset safety conditions, then the path length and cumulative curvature of the first parking path are determined. The path operation score of the first parking path is determined based on the path length and the cumulative curvature of the path. In response to the existence of a first parking path that meets the preset safety conditions, a target parking path is determined based on the path operation score of the first parking path that meets the preset safety conditions, and the vehicle is controlled to park according to the target parking path.

8. The method according to claim 1, characterized in that, The parking reference information includes map boundary and obstacle information. Based on the parking reference information and the vehicle kinematics model, a second parking path is searched from the current pose to the final parking pose, including: An occupation grid is constructed based on the map boundary and the obstacle information, and passable areas are marked in the occupation grid; Using the current pose as the node to be searched, and based on the vehicle kinematics model, the next optimal node of the node to be searched is searched in the passable area. The vehicle kinematics model describes the relationship between the vehicle's pose and the control input, and is used to determine the next node based on the node to be searched and the control input. The next optimal node is taken as the new node to be searched, and the step of searching for the next optimal node of the node to be searched in the passable area based on the vehicle kinematics model is returned until the final parking pose is reached, so as to obtain the second parking path from the current pose to the final parking pose.

9. The method according to claim 1, characterized in that, Before controlling the vehicle to park based on the second parking path, the method further includes: Obtain the preset path length and preset rate of curvature change; The second parking path is adjusted and updated with the goal of ensuring that the path length of the second parking path does not exceed the preset path length and that the rate of curvature change of the second parking path does not exceed the preset rate of curvature change.

10. An electronic device, characterized in that, The electronic device includes: Processor and memory; The processor executes the steps of the parking method as described in any one of claims 1 to 9 by calling programs or instructions stored in the memory.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program or instructions that cause a computer to perform the steps of the parking method as described in any one of claims 1 to 9.

Images