A parking path planning method and device, a vehicle and a readable storage medium

Abstract

The embodiment of the application provides a parking path planning method, device, vehicle and readable storage medium, the parking path planning method obtains the starting pose and target pose of a vehicle to be parked, converts a grid map into a cost map based on the starting pose and target pose through an A-star algorithm, plans an extended path of the vehicle to be parked in the grid map based on the cost map through a hybrid A-star algorithm, plans an RS path of the vehicle to be parked by using an RS geometric algorithm, combines the extended path and the RS path, and performs collision detection on the combined path to obtain a planned parking path. The application can avoid the problems of scene limitation and excessive dependence on environmental information in parking, is suitable for all parking scenes, and thus can improve the user's sense of experience.

Description

Technical Field

[0001] This invention relates to the field of parking route planning, and more specifically, to a parking route planning method, apparatus, vehicle, and readable storage medium. Background Technology

[0002] During automatic parking, the perception system uses cameras or ultrasonic radar to identify parking spaces and obstacles, and then transmits the identification results to the planning and decision-making system. The planning and decision-making system first determines whether the parking space meets the requirements. If it does, it plans a parking path from the starting position and orientation to the parking space. Existing automatic parking path planning generally uses a geometric method consisting of circular arcs and / or straight lines. However, this method is heavily dependent on environmental information, such as aisle width and distance from the parking starting point to the target parking space, and lacks universality. Summary of the Invention

[0003] The purpose of this invention is to provide a parking route planning method, apparatus, vehicle, and readable storage medium.

[0004] In a first aspect, the present invention provides a parking path planning method applied to vehicles waiting to park, the method comprising:

[0005] Obtain the initial pose and target pose of the vehicle waiting to be parked;

[0006] Based on the initial pose and the target pose, the raster map is converted into a cost map using the A* algorithm;

[0007] Based on the cost map, the extended path of the waiting vehicle is planned in the grid map using the hybrid A* algorithm;

[0008] The RS geometric algorithm is used to plan the RS path of the vehicle waiting to be parked. The extended path is combined with the RS path, and collision detection is performed on the combined path to obtain the planned parking path.

[0009] In an optional implementation, before acquiring the initial pose and target pose of the vehicle to be parked, the method further includes:

[0010] Obtain environmental information from a preset map and the area around the parked vehicle;

[0011] The preset map is converted into a raster map based on the preset raster resolution and the environmental information.

[0012] In an optional implementation, the step of planning the extended path of the waiting vehicle in the grid map based on the cost map using a hybrid A* algorithm includes:

[0013] Using the initial pose of the vehicle waiting to park as the initial parent node, multiple candidate expansion nodes are determined in the grid map according to preset distance, preset angle and preset step size;

[0014] The collision detection determines whether a collision occurs when the vehicle waiting to be parked travels to each candidate expansion node.

[0015] If a collision occurs at the candidate expansion node, the search for the expansion path is terminated.

[0016] If no collision occurs at the candidate expansion node, the candidate expansion node is determined to be an expandable node. The total generation value of each expandable node is calculated according to the cost strategy and the cost map, and the expandable node with the smallest total generation value among all expandable nodes is determined as the latest expansion node.

[0017] Determine whether the latest extended node meets the preset termination condition;

[0018] When the latest extended node does not meet the preset termination condition, the latest extended node will be taken as the parent node, and the process of determining multiple extended nodes according to the preset distance, preset angle and preset step size will be re-executed.

[0019] When the latest extended node meets the preset termination condition, the search for the extended path is terminated and the latest extended node and its series of parent nodes are output as the extended path.

[0020] In an optional implementation, the process of performing collision detection on the vehicle waiting to park includes:

[0021] Determine the four boundary points of the vehicle to be parked in the vehicle coordinate system;

[0022] Convert the four boundary points in the vehicle coordinate system to four boundary points in the world coordinate system;

[0023] The rectangular frame constituting the vehicle waiting to be parked is determined based on the four boundary points in the world coordinate system.

[0024] Determine whether there is an obstacle line intersecting with the rectangle based on the obstacle lines on the grid map and the rectangle.

[0025] In an optional implementation, calculating the total generation value of the scalable node based on the cost strategy and cost map includes:

[0026] The dissipation cost of the scalable node is determined based on a cost strategy;

[0027] The heuristic cost is determined based on the cost map, and the total generation value of the scalable node is determined based on the dissipation cost and the heuristic cost.

[0028] In an optional implementation, determining whether the latest extended node meets a preset termination condition includes:

[0029] Determine if the list containing candidate expansion nodes is empty;

[0030] When the list is empty, it is determined that the latest extended node satisfies the preset termination condition;

[0031] When the candidate expansion node exists in the list, it is determined that the latest expansion node does not meet the preset termination condition.

[0032] In an optional implementation, the preset map is either a high-precision map or a positioning and map building tool.

[0033] Secondly, the present invention provides a parking path planning device for use with vehicles waiting to park, the device comprising:

[0034] The acquisition module is used to acquire the initial pose and target pose of the vehicle waiting to be parked.

[0035] The cost module is used to convert the grid map into a cost map based on the initial pose and the target pose using the A* algorithm;

[0036] An extended path module is used to plan an extended path for the waiting vehicle in the grid map based on the cost map using a hybrid A* algorithm;

[0037] The RS path module is used to plan the RS path of the vehicle waiting to be parked using the RS geometric algorithm, combine the extended path with the RS path, and perform collision detection on the combined path to obtain the planned parking path.

[0038] Thirdly, the present invention provides a vehicle including a memory and a processor, the memory storing a computer program that executes the parking path planning method when the computer program is run on the processor.

[0039] Fourthly, the present invention provides a readable storage medium storing a computer program that executes the parking path planning method when run on a processor.

[0040] The beneficial effects of the embodiments of the present invention are:

[0041] This application provides a parking path planning method. The method obtains the initial and target poses of the vehicle to be parked. Based on these poses, it converts a grid map into a cost map using the A* algorithm. Then, based on the cost map, it plans an extended path for the vehicle within the grid map using a hybrid A* algorithm. Finally, it plans an RS path for the vehicle using the RS geometric algorithm. The extended path and RS path are combined, and collision detection is performed on the combined path to obtain the planned parking path. This method avoids the limitations of parking scenarios and excessive reliance on environmental information, making it applicable to all parking scenarios and thus improving the user experience.

[0042] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0043] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope of protection of the present invention. In the various drawings, similar components are numbered similarly.

[0044] Figure 1 This paper illustrates a first flowchart of a parking path planning method proposed in an embodiment of this application.

[0045] Figure 2 This paper illustrates a second flowchart of a parking path planning method proposed in an embodiment of this application.

[0046] Figure 3 The flowchart of a parking path planning method with hybrid A* algorithm proposed in an embodiment of this application is shown.

[0047] Figure 4 This paper illustrates a flowchart of collision detection in a parking path planning method proposed in an embodiment of this application.

[0048] Figure 5 This paper illustrates a flowchart of a parking path planning method for calculating the total cost value according to an embodiment of this application.

[0049] Figure 6 This paper illustrates a flowchart of a parking path planning method proposed in an embodiment of this application to determine whether a preset termination condition is met.

[0050] Figure 7 A schematic diagram of a parking path planning device provided in an embodiment of this application is shown.

[0051] Explanation of key component symbols:

[0052] 10-Parking route planning device; 11-Acquisition module; 12-Cost module; 13-Extended route module; 14-RS route module. Detailed Implementation

[0053] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0054] The components of the embodiments of the invention described and illustrated herein can typically be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0055] In the following, the terms “comprising,” “having,” and their cognates, which may be used in various embodiments of the invention, are intended only to indicate a particular feature, number, step, operation, element, component, or combination thereof, and should not be construed as excluding, firstly, the presence of one or more other features, numbers, steps, operations, elements, components, or combinations thereof, or adding the possibility of one or more features, numbers, steps, operations, elements, components, or combinations thereof.

[0056] Furthermore, the terms "first," "second," and "third" are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.

[0057] Unless otherwise specified, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which the various embodiments of the invention pertain. Terms (such as those defined in commonly used dictionaries) shall be interpreted as having the same meaning as in their contextual meaning in the relevant technical field and shall not be interpreted as having an idealized or overly formal meaning, unless clearly defined in the various embodiments of the invention.

[0058] Example 1

[0059] Please refer to Figure 1 This application proposes a parking path planning method for use in vehicles. This method can plan the parking path of a vehicle to be parked, thereby enabling automatic parking. Exemplarily, the parking path planning method includes steps S100 to S300.

[0060] Step S100: Obtain the initial pose and target pose of the vehicle to be parked.

[0061] Understandably, the initial pose of the vehicle waiting to park is used as the starting point. When the user sends an automatic parking command via the human-machine interface or voice, the vehicle will initiate automatic parking upon receiving the command. After initiating automatic parking, the vehicle will travel at low speed to search for parking spaces. Once it finds available parking spaces around the vehicle, the grid map will display the corresponding space number and pose information. At this point, the vehicle will stop, and the user will select the target parking space. Typically, the vehicle will stop 4-5 meters from the target parking space. After the user selects the target parking space, the pose of that target space will be determined as the target pose. Here, pose represents position and attitude. The starting point of the vehicle is the position where it stops after finding a parking space. This position can be automatically determined using the vehicle's GPS module, and the pose corresponding to this starting point is the initial pose of the vehicle.

[0062] In one implementation, such as Figure 2 As shown, steps S500 to S600 are included before step S100.

[0063] Step S500: Obtain environmental information on the preset map and around the vehicle waiting to be parked.

[0064] In this embodiment of the application, the preset map is a map scanned by the vehicle waiting to park, which can be a high-precision map (Map for Highly Automated Driving, HAD Map) or a SLAM (Simultaneous Localization and Mapping, also known as CML, Concurrent Mapping and Localization) map.

[0065] In this embodiment, environmental information surrounding the vehicle waiting to park can be obtained through a preset map, or through ultrasonic sensors, surround-view cameras, or lidar installed on the vehicle. This means obtaining the location information of the surrounding environment in world coordinates. Environmental information includes, but is not limited to, lane lines, the location information of static obstacles, and parking space lines. Static obstacles include vehicles, parking locks, and traffic cones. Exemplarily, when the preset map is a high-precision map, it can provide information such as the geographical coordinates of lanes, pedestrian crossings, traffic lights, and lane lines.

[0066] Step S600: Convert the preset map into a raster map based on the preset raster resolution and environmental information.

[0067] Understandably, appropriate obstacle lines will be set based on the environmental information surrounding the vehicle waiting to park. After the user selects a target parking space, the scanned preset map will be converted into a raster map based on the preset grid resolution and environmental information. This involves converting world coordinates combined with environmental information into raster coordinates. The preset grid resolution can be set according to actual needs. The grid resolution affects the number of grid cells occupied by obstacles. Therefore, provided the preset calculation time is met, a lower preset grid resolution can be selected to improve the parking success rate of the vehicle waiting to park.

[0068] In other words, the preset map is divided into grids of preset size to obtain an initial grid map. Obstacle lines are then set on the initial grid map based on environmental information, i.e., by setting the grids with corresponding numbers to obtain the corresponding grid map. For example, the grid value corresponding to an obstacle grid can be set to 1, and the grid values ​​of other grids can be set to 0, thus forming the obstacle lines in the grid map. The lane line from the lane where the waiting vehicle enters the selected target parking space cannot be used as an obstacle line. Instead, it connects static obstacles, passageway lane lines, and parking space lines present in other environmental information to form the obstacle lines on the grid map, thus constituting the drivable area for the waiting vehicle.

[0069] Step S200: Based on the initial pose and target pose, convert the grid map into a cost map using the A* algorithm.

[0070] In this embodiment, after determining the initial pose and target pose of the vehicle waiting to park, the initial pose of the vehicle is used as the starting point, and the target pose is used as the target grid in the grid map. The corresponding obstacle grid in the cost map is filled with inf (infinite) to indicate that the point is unreachable. The cost value from each grid in the grid map (excluding obstacle grids) to the target grid is calculated using the A* algorithm, and the calculated cost value is filled into the corresponding grid to obtain the corresponding cost map. This cost map represents the cost value from each grid to the target grid. Based on this cost map, the direction in which the vehicle waiting to park reaches the target pose fastest is determined; that is, this direction can serve as a global reference path and guide the direction of the extended path for the hybrid A* algorithm.

[0071] For example, if the grid value corresponding to a grid in the grid map is 1, that is, the grid is an obstacle grid, then inf will be filled into the grid in the cost map; when the grid value corresponding to a grid in the grid map is 0, then the cost value from the grid to the target grid will be calculated by the A* algorithm, and the obtained cost value will be filled into the corresponding grid in the cost map.

[0072] Step S300: Based on the cost map, plan the extended path of the waiting vehicles in the grid map using the hybrid A* algorithm.

[0073] Understandably, after determining the approximate direction of the parking path of the waiting vehicle through the cost map, the hybrid A* algorithm will be used to plan an extended path on the grid map that allows the waiting vehicle to avoid obstacles.

[0074] In one implementation, such as Figure 3 As shown, step S300 also includes executing sub-steps S310 to S300 with the initial pose of the vehicle waiting to be parked as the initial parent node.

[0075] Sub-step S310: Determine multiple candidate expansion nodes in the grid map according to preset distance, preset angle and preset step size.

[0076] In this embodiment, the initial pose of the vehicle waiting to park, i.e., the starting point, is used as the initial parent node. Centered on this initial parent node, multiple candidate expansion nodes are determined in the grid map according to preset distances, preset angles, and preset step sizes. The preset angle refers to the front wheel deflection angle of the vehicle waiting to park. The preset step size, preset distance, and preset angle are set according to the actual scenario.

[0077] Sub-step S320: Determine whether a collision occurs when the vehicle waiting to park travels to each candidate expansion node through collision detection.

[0078] After determining multiple candidate expansion nodes corresponding to the initial parent node, collision detection is performed on each candidate expansion node based on its pose to determine whether a collision will occur when the vehicle to be parked reaches that candidate expansion node. If the collision test shows that the vehicle to be parked will collide with the candidate expansion node, it means that the hybrid A* algorithm has no solution, and sub-step S370 is executed; if the collision test shows that the vehicle to be parked will not collide with the candidate expansion node, then sub-step S330 is executed.

[0079] In one implementation, such as Figure 4 As shown, the process of collision detection for a vehicle to be parked includes sub-steps S321 to S324.

[0080] Sub-step S321: Determine the four boundary points of the vehicle to be parked in the vehicle coordinate system.

[0081] In this application, sensors mounted on the vehicle will determine the four boundary points of the vehicle waiting to park in the vehicle coordinate system, specifically calculating the positions corresponding to the left front, right front, left rear, and right rear of the vehicle. The vehicle coordinate system is a special moving coordinate system used to describe the motion of a car; its origin coincides with the center of mass. When the vehicle is stationary on a level surface, the X-axis is parallel to the ground and points forward, the Z-axis passes through the car's center of mass and points upward, and the Y-axis points to the driver's left.

[0082] Sub-step S322: Convert the four boundary points in the vehicle coordinate system to four boundary points in the world coordinate system.

[0083] The positions corresponding to the left front, right front, left rear, and right rear of the vehicle waiting to park in the vehicle coordinate system are transformed to the world coordinate system to obtain the positioning information of the vehicle waiting to park in the world coordinate system. The rear axle center of the vehicle in its initial pose is taken as the origin, and the vehicle's front-facing angle is taken as the positive direction.

[0084] Sub-step S323: Determine the rectangular frame that constitutes the parking vehicle based on the four boundary points in the world coordinate system.

[0085] By using the four boundary points of the vehicle waiting to park in the world coordinate system—namely, the positions of the left front, right front, left rear, and right rear—four line segments will be formed to create the rectangular frame corresponding to the vehicle.

[0086] Sub-step S324: Determine whether there are any obstacle lines intersecting with the rectangles based on the obstacle lines and rectangles on the grid map.

[0087] In this embodiment, the obstacle lines set in the grid map are traversed, and it is determined whether the four line segments of the rectangle corresponding to the parking vehicle intersect with the obstacle lines when the parking vehicle travels to each candidate expansion node. If there is an obstacle line where the four line segments corresponding to the parking vehicle intersect with the set obstacle lines, it indicates that the parking vehicle will collide with obstacles in the surrounding environment when it travels to the candidate expansion node; if the four line segments corresponding to the parking vehicle do not intersect with any set obstacle lines, it indicates that the parking vehicle will not collide when it travels to the candidate expansion node.

[0088] This implementation uses a rectangular frame composed of the vehicle's length and width for collision detection, completely covering the entire vehicle body, ensuring safety without blind spots.

[0089] Sub-step S330: Determine candidate expansion nodes as expandable nodes, calculate the total generation value of each expandable node according to the cost strategy and cost map, and determine the expandable node with the smallest total generation value among all expandable nodes as the latest expansion node.

[0090] When a vehicle waiting to park does not collide with a candidate expansion node, the candidate expansion node is designated as an expandable node. The total cost value of each expandable node is calculated based on the cost map and cost strategy. The expandable node with the lowest total cost value among all expandable nodes is designated as the latest expansion node.

[0091] In one implementation, such as Figure 5 As shown, the total value of the scalable node is calculated based on the cost strategy and cost map, including sub-steps S331 to S332.

[0092] Sub-step S331: Determine the dissipation cost of scalable nodes based on the cost strategy.

[0093] In this embodiment, the dissipation cost G corresponding to the scalable node will be calculated according to the cost strategy, which includes the cost of fewer gear shifts, the cost of no sharp turns, and the cost of discontinuous turns.

[0094] The shift cost is calculated using the formula cost1 = f + config.switch_cost * distance, where f represents the cost calculated by the scalable node according to the A* algorithm, config.switch_cost represents the shift cost gain, and distance represents the distance to the next node. When fewer shifts are needed, config.switch_cost is set to a larger value, resulting in a higher calculated shift cost, and the scalable node will not be selected. The turning cost is calculated using the formula cost2 = f + config.steer_cost * abs(delta), where cost represents the turning cost gain, config.steer_cost represents the steering angle, and delta represents the steering angle. When there are no sharp turns, config.steer_cost is set to a larger value. The continuous turning cost is calculated using the formula cost3 = f + config.steer_continue_cost * abs(delta - parent.delta), where config.steer_continue_cost represents the continuous turning cost gain, and delta - parent.delta represents the turning change amount. When there are no continuous turns, config.steer_continue_cost is set to a larger value.

[0095] Sub-step S332: Determine the heuristic cost based on the cost map, and determine the total value of the scalable node based on the dissipation cost and the heuristic cost.

[0096] Understandably, the cost map obtained through the A* algorithm is used as the heuristic cost H. The cost map is then used to calculate the heuristic cost H, specifically using the formula H = costmap(node.idx, node.idy). The cost map represents the approximate direction of the parking paths of the searched vehicles, thus avoiding the problem of incorrect orientation expansion during node expansion using a hybrid A* algorithm, and accelerating the search process. After determining the dissipative cost and heuristic cost, the total value corresponding to each expandable node is calculated using the appropriate formula.

[0097] Specifically, the total generation value corresponding to each scalable node is calculated using the following formula:

[0098] F = G + H, where F represents the total generation value corresponding to the scalable node, G represents the dissipation cost, and H represents the heuristic cost.

[0099] Sub-step S340: Determine whether the latest expanded node meets the preset termination conditions.

[0100] Determine whether the latest extended node meets the preset termination condition. If the latest extended node does not meet the preset termination condition, execute sub-step S350. If the latest extended node meets the preset termination condition, execute sub-step S360.

[0101] In one implementation, such as Figure 6 As shown, determining whether the latest expanded node meets the preset termination conditions includes sub-steps S341 to S343.

[0102] Sub-step S341: Determine whether the list storing candidate expansion nodes is empty.

[0103] In this embodiment, an open list and a close list are pre-generated. The open list stores candidate expansion nodes and is initially empty. The starting point is added to the open list, and the starting point in the open list is used as the center point to expand outwards to 8 child nodes. The path from the reference point to each child node is calculated sequentially, and the shortest child node is determined and written to the open list. When there are no candidate expansion nodes in the open list, it means that the list is empty, and sub-step S342 is executed. When the list is not empty, sub-step S343 is executed.

[0104] Sub-step S342: Determine if the latest extended node meets the preset termination condition.

[0105] If no expandable node exists in the list of candidate expandable nodes, the latest expandable node is considered to meet the preset termination condition.

[0106] Sub-step S343: Determine that the latest extended node does not meet the preset termination condition.

[0107] If there are candidate expansion nodes in the list, the latest expansion node is considered not to meet the preset termination condition.

[0108] Sub-step S350: Take the latest expanded node as the parent node and re-execute sub-step S310.

[0109] Sub-step S360: Output the latest extended node and its series of parent nodes as the extended path, and execute sub-step S370.

[0110] When the latest expanded node meets the preset termination condition, the latest expanded node and its series of parent nodes will be output as expanded path points.

[0111] Sub-step S370: Terminate the search for extended paths.

[0112] Step S400: Use the RS geometric algorithm to plan the RS path for the vehicle waiting to park, combine the extended path with the RS path, and perform collision detection on the combined path to obtain the planned parking path.

[0113] In this embodiment of the application, after determining the extended path of the vehicle waiting to be parked, RS path planning will be performed based on the endpoint of the extended path of the vehicle waiting to be parked. In other words, when planning the RS path of the vehicle waiting to be parked through the RS geometric algorithm, the last latest extended node of the extended path will be used as the starting point of the RS path, and the target pose will be used as the endpoint for RS path planning.

[0114] After obtaining the RS path planned using the RS geometric algorithm, this RS path is combined with the extended path obtained using the hybrid A* algorithm. Collision detection is then performed on the combined path to determine if the vehicle's bounding box intersects with any set obstacle lines as it travels along the combined path. The path after collision detection is then used as the planned parking path. This ensures that the planned path direction does not deviate from the target direction when the vehicle reaches the target parking space. The RS geometric algorithm is a route planning method based on geometric algorithms, capable of quickly planning a path from the starting point to the ending point (i.e., the target point). The path planned using the RS geometric algorithm is an RS curve, short for Reeds-Shepp curve. This curve not only ensures that the vehicle reaches the destination but also that the vehicle's angle reaches the expected angle at the destination.

[0115] This application avoids the problems of parking scenario limitations and over-reliance on environmental information, and is applicable to all parking scenarios, thereby improving the user experience.

[0116] The parking path planning method based on the above embodiments Figure 7A schematic diagram of a parking path planning device 10 according to an embodiment of this application is shown, which is applied to a vehicle waiting to park. The parking path planning device 10 includes:

[0117] The acquisition module 11 is used to acquire the initial pose and target pose of the vehicle waiting to be parked;

[0118] Cost module 12 is used to convert the grid map into a cost map based on the initial pose and the target pose using the A* algorithm;

[0119] Extended path module 13 is used to plan extended paths for vehicles waiting to park in a grid map based on the cost map using a hybrid A* algorithm;

[0120] RS path module 14 is used to plan the RS path of the vehicle waiting to be parked using the RS geometric algorithm, combine the extended path with the RS path, and perform collision detection on the combined path to obtain the planned parking path.

[0121] The parking path planning device 10 in this embodiment is used to execute the parking path planning method of the above embodiment. The implementation schemes and beneficial effects involved in the above embodiments are also applicable in this embodiment, and will not be repeated here.

[0122] This application embodiment also provides a vehicle, which includes a vehicle body and a parking path planning device 10. The parking path planning device is disposed on the vehicle body and is used to plan the parking path of the vehicle to be parked.

[0123] This application also provides a computer-readable storage medium storing a computer program that, when executed on a processor, implements the above-described parking path planning method.

[0124] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that, as an alternative implementation, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0125] In addition, the functional modules or units in the various embodiments of the present invention can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0126] If the functionality is implemented as a software module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a smartphone, personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0127] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A parking path planning method, characterized in that, Applied to vehicles waiting to be parked, the method includes: Obtain the initial pose and target pose of the vehicle waiting to be parked; Based on the initial pose and the target pose, the raster map is converted into a cost map using the A* algorithm; Based on the cost map, the extended path of the waiting vehicle is planned in the grid map using the hybrid A* algorithm; The RS geometric algorithm is used to plan the RS path of the vehicle waiting to be parked. The extended path is combined with the RS path, and collision detection is performed on the combined path to obtain the planned parking path. The step of planning the expansion path of the waiting vehicle in the grid map using the hybrid A* algorithm based on the cost map includes: taking the initial pose of the waiting vehicle as the initial parent node, determining multiple candidate expansion nodes in the grid map according to a preset distance, preset angle, and preset step size; determining whether a collision occurs when the waiting vehicle travels to each candidate expansion node through collision detection; if no collision occurs at the candidate expansion node, determining the candidate expansion node as an expandable node; calculating the total cost value of each expandable node according to the cost strategy and the cost map; and determining the expandable node with the smallest total cost value among all expandable nodes as the latest expansion node. After determining the scalable node with the lowest total value among all scalable nodes as the latest scalable node, the process further includes: determining whether the latest scalable node meets a preset termination condition; if the latest scalable node does not meet the preset termination condition, the latest scalable node will be used as the parent node, and the process of determining multiple scalable nodes according to a preset distance, preset angle, and preset step size will be re-executed; if the latest scalable node meets the preset termination condition, the search for the scalable path will be terminated, and the latest scalable node and its series of parent nodes will be output as the scalable path. The process of collision detection for the vehicle waiting to be parked includes: determining four boundary points of the vehicle waiting to be parked in the vehicle coordinate system; converting the four boundary points in the vehicle coordinate system into four boundary points in the world coordinate system; determining the rectangular frame constituting the vehicle waiting to be parked based on the four boundary points in the world coordinate system; and determining whether there are any obstacle lines intersecting the rectangular frame based on the obstacle lines on the grid map and the rectangular frame. The step of calculating the total generation value of the scalable node based on the cost strategy and the cost map includes: determining the dissipation cost of the scalable node based on the cost strategy; wherein the cost strategy includes cost of fewer gear shifts, cost of no sharp turns, and cost of discontinuous turns; determining the heuristic cost based on the cost map; and determining the total generation value of the scalable node based on the dissipation cost and the heuristic cost. The formula for calculating the total cost is: F = G + H, where F represents the total generation value corresponding to the scalable node, G represents the dissipation cost, and H represents the heuristic cost.

2. The parking path planning method according to claim 1, characterized in that, Before acquiring the initial and target poses of the vehicle to be parked, the method further includes: Obtain environmental information from a preset map and the area around the parked vehicle; The preset map is converted into a raster map based on the preset raster resolution and the environmental information.

3. The parking path planning method according to claim 1, characterized in that, Determining whether the latest extended node meets the preset termination condition includes: Determine if the list containing candidate expansion nodes is empty; When the list is empty, it is determined that the latest extended node satisfies the preset termination condition; When the candidate expansion node exists in the list, it is determined that the latest expansion node does not meet the preset termination condition.

4. The parking path planning method according to claim 2, characterized in that, The preset map can be either a high-precision map or a positioning and map construction map.

5. A parking path planning device, characterized in that, Applied to vehicles waiting to park, the device includes: The acquisition module is used to acquire the initial pose and target pose of the vehicle waiting to be parked. The cost module is used to convert the grid map into a cost map based on the initial pose and the target pose using the A* algorithm; An extended path module is used to plan an extended path for the waiting vehicle in the grid map based on the cost map using a hybrid A* algorithm; The RS path module is used to plan the RS path of the vehicle waiting to be parked using the RS geometric algorithm, combine the extended path with the RS path, and perform collision detection on the combined path to obtain the planned parking path. The step of planning the expansion path of the waiting vehicle in the grid map using the hybrid A* algorithm based on the cost map includes: taking the initial pose of the waiting vehicle as the initial parent node, determining multiple candidate expansion nodes in the grid map according to a preset distance, preset angle, and preset step size; determining whether a collision occurs when the waiting vehicle travels to each candidate expansion node through collision detection; if no collision occurs at the candidate expansion node, determining the candidate expansion node as an expandable node; calculating the total cost value of each expandable node according to the cost strategy and the cost map; and determining the expandable node with the smallest total cost value among all expandable nodes as the latest expansion node. After determining the scalable node with the lowest total value among all scalable nodes as the latest scalable node, the process further includes: determining whether the latest scalable node meets a preset termination condition; if the latest scalable node does not meet the preset termination condition, the latest scalable node will be used as the parent node, and the process of determining multiple scalable nodes according to a preset distance, preset angle, and preset step size will be re-executed; if the latest scalable node meets the preset termination condition, the search for the scalable path will be terminated, and the latest scalable node and its series of parent nodes will be output as the scalable path. The process of collision detection for the vehicle waiting to be parked includes: determining four boundary points of the vehicle waiting to be parked in the vehicle coordinate system; converting the four boundary points in the vehicle coordinate system into four boundary points in the world coordinate system; determining the rectangular frame constituting the vehicle waiting to be parked based on the four boundary points in the world coordinate system; and determining whether there are any obstacle lines intersecting the rectangular frame based on the obstacle lines on the grid map and the rectangular frame. The step of calculating the total generation value of the scalable node based on the cost strategy and the cost map includes: determining the dissipation cost of the scalable node based on the cost strategy; wherein the cost strategy includes cost of fewer gear shifts, cost of no sharp turns, and cost of discontinuous turns; determining the heuristic cost based on the cost map; and determining the total generation value of the scalable node based on the dissipation cost and the heuristic cost. The formula for calculating the total cost is: F = G + H, where F represents the total generation value corresponding to the scalable node, G represents the dissipation cost, and H represents the heuristic cost.

6. A vehicle, characterized in that, It includes a memory and a processor, the memory storing a computer program that, when run on the processor, executes the parking path planning method according to any one of claims 1 to 4.

7. A readable storage medium, characterized in that, It stores a computer program that, when run on a processor, executes the parking path planning method according to any one of claims 1 to 4.

Images