Parking path planning method and device, electronic equipment and medium

Abstract

The embodiment of the invention discloses a parking path planning method and device, electronic equipment and a medium, and belongs to the technical field of automatic driving. The method comprises the steps that a parking grid map is generated according to vehicle information of a to-be-parked vehicle and environment information of the environment where the to-be-parked vehicle is located, and the parking grid map is used for indicating obstacle information and parking space information within a preset range; based on the parking grid map, at least one first alternative parking path is determined according to the first path cost, and the first alternative parking path is a path with the to-be-parked vehicle as a starting point and the free parking space as an ending point; for any first alternative parking path, optimizing at least part of paths in the first alternative parking paths according to the second path cost to obtain second alternative parking paths corresponding to the first alternative parking paths; and selecting a target parking path from the second alternative parking paths corresponding to the first alternative parking paths. According to the method, a more reasonable and convenient parking path can be obtained, and the parking success rate can be improved.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and in particular to a parking path planning method and a parking path planning device, an electronic device, and a computer-readable medium. Background Art

[0002] When a user activates an autonomous vehicle system, the path planning decision system must plan the required path within a specified timeframe to assist driving. However, the limited resources of the onboard computing platform and the high algorithm overhead affect the success rate of path planning. Therefore, successfully and efficiently implementing path planning for autonomous vehicles within the constraints of available resources has become a pressing issue. Summary of the Invention

[0003] To this end, the present disclosure provides a parking path planning method and a parking path planning device, an electronic device, and a computer-readable medium.

[0004] To achieve the above-mentioned objective, a first aspect of the present disclosure provides a parking path planning method, the method comprising: generating a parking grid map based on vehicle information of a vehicle to be parked and environmental information of an environment in which the vehicle to be parked is located, the parking grid map being used to indicate obstacle information and parking space information within a preset range;

[0005] determining, based on the parking grid map and according to a first path cost, at least one first alternative parking path, the first alternative parking path being a path starting from the vehicle to be parked and ending at an empty parking space;

[0006] For any of the first alternative parking paths, optimizing at least a portion of the first alternative parking paths according to a second path cost to obtain a second alternative parking path corresponding to the first alternative parking path;

[0007] A target parking path is selected from the second candidate parking paths corresponding to each of the first candidate parking paths.

[0008] Optionally, the vehicle information of the vehicle to be parked includes position information of the vehicle to be parked;

[0009] Generating a parking grid map based on vehicle information of the vehicle to be parked and environmental information of an environment in which the vehicle to be parked is located includes:

[0010] Performing target object recognition based on environmental information of the environment in which the vehicle to be parked is located to obtain target object recognition information, wherein the target objects include obstacles and vacant parking spaces;

[0011] The parking grid map is generated based on preset map parameters and according to the position information of the vehicle to be parked and the target object recognition information.

[0012] Optionally, the environmental information includes a plurality of environmental images;

[0013] The step of identifying a target object based on environmental information of an environment in which the vehicle to be parked is located to obtain target object identification information includes:

[0014] Inputting the plurality of environment images into a preset target detection model to obtain target object recognition information;

[0015] The target detection model is used to mark a target frame in the environment image, and the target frame is used to select a target object in the environment image.

[0016] Optionally, the target object recognition information includes obstacle distribution information and position information of the vacant parking space;

[0017] The generating of the parking grid map based on preset map parameters and according to the position information of the vehicle to be parked and the target object recognition information includes:

[0018] Based on the preset map parameters, generating an initial grid map;

[0019] According to the posture information of the vehicle to be parked, the obstacle distribution information and the posture information of the vacant parking spaces, the vehicle to be parked, the obstacles and the vacant parking spaces are mapped into the initial grid map to obtain the parking grid map.

[0020] Optionally, the first path cost includes an actual cost and a heuristic cost of the corresponding path, and the second path cost includes the actual cost of the corresponding path.

[0021] Optionally, determining at least one first alternative parking path based on the parking grid map and according to a first path cost includes:

[0022] For the i-th processing step, a plurality of expansion nodes corresponding to the i-th parent node are determined in the parking grid map according to a preset expansion parameter, where i is an integer greater than or equal to 1, and the first parent node is a mapping node of the vehicle to be parked in the parking grid map;

[0023] Determine the first path cost corresponding to each extended node;

[0024] selecting an updated extension node from the plurality of extension nodes according to the first path cost corresponding to each of the extension nodes;

[0025] In a case where the updated extended node does not belong to the target node, performing the (i+1)th processing process with the updated extended node as the (i+1)th parent node, the target node being a mapping node of the vacant parking space in the parking grid map;

[0026] In a case where the updated extended node belongs to the target node, the first candidate parking path is determined according to the 1st to i-th parent nodes and the updated extended node.

[0027] Optionally, optimizing at least a portion of the first candidate parking path according to the second path cost to obtain a second candidate parking path corresponding to the first candidate parking path includes:

[0028] In any optimization process of the first candidate parking path, a portion of the first candidate parking path is selected to obtain a first sub-path;

[0029] Taking the first node of the first subpath as a new starting point and the last node of the first subpath as a new end point, determining a second subpath according to the second path cost;

[0030] When the path length of the second sub-path is shorter than the path length of the first sub-path, the first sub-path in the first candidate parking path is replaced by the second sub-path.

[0031] A second aspect of the present disclosure provides a parking path planning device, the device comprising: a generating module for generating a parking grid map based on vehicle information of a vehicle to be parked and environmental information of an environment in which the vehicle to be parked is located, the parking grid map being used to indicate obstacle information and parking space information within a preset range;

[0032] a determining module configured to determine, based on the parking grid map and according to a first path cost, at least one first alternative parking path, the first alternative parking path being a path starting from the vehicle to be parked and ending at an empty parking space;

[0033] an optimization module configured to optimize, for any of the first alternative parking paths, at least a portion of the first alternative parking paths according to a second path cost, to obtain a second alternative parking path corresponding to the first alternative parking path;

[0034] A selection module is configured to select a target parking path from the second candidate parking paths corresponding to each of the first candidate parking paths.

[0035] A third aspect of the present disclosure provides an electronic device, the electronic device comprising: one or more processors;

[0036] a storage device storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the parking path planning method according to any one of the embodiments of the present disclosure;

[0037] One or more I / O interfaces are connected between the processor and the memory and configured to implement information interaction between the processor and the memory.

[0038] A fourth aspect of the present disclosure provides a computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the parking path planning method described in any one of the embodiments of the present disclosure.

[0039] The present disclosure has the following advantages: a parking grid map is generated based on vehicle information of a vehicle to be parked and environmental information of the environment in which the vehicle is parked. The parking grid map indicates obstacle information and parking space information within a preset range; at least one first alternative parking path is determined based on the parking grid map according to a first path cost, where the first alternative parking path is a path starting from the vehicle to be parked and ending at an available parking space; for any first alternative parking path, at least a portion of the first alternative parking path is optimized based on a second path cost to obtain a second alternative parking path corresponding to the first alternative parking path; and a target parking path is selected from the second alternative parking paths corresponding to each first alternative parking path. Thus, the present disclosure embodiment can, on the one hand, determine at least one first alternative parking path based on the first path cost, and on the other hand, optimize the first alternative parking path using the second path cost to obtain a more reasonable and convenient parking path, thereby improving the parking success rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings are used to provide further understanding of the present disclosure and constitute a part of the specification. Together with the following detailed description, they are used to explain the present disclosure, but do not constitute a limitation of the present disclosure.

[0041] Figure 1 A flowchart of a parking path planning method provided in an embodiment of the present disclosure.

[0042] Figure 2 A flowchart of a parking path planning method provided in an embodiment of the present disclosure.

[0043] Figure 3 A block diagram of a parking path planning device provided in an embodiment of the present disclosure.

[0044] Figure 4 A block diagram of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0045] The following describes the specific embodiments of the present disclosure in detail with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to illustrate and explain the present disclosure and are not intended to limit the present disclosure.

[0046] As used in this disclosure, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0047] The terms used in the present disclosure are only used to describe specific embodiments and are not intended to limit the present disclosure.As used in the present disclosure, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

[0048] When the terms “comprising” and / or “made of…” are used in the present disclosure, it specifies the existence of the stated features, integers, steps, operations, elements and / or components, but does not preclude the existence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0049] The embodiments of the present disclosure may be described with reference to plan views and / or cross-sectional views by way of ideal schematic views of the present disclosure. Therefore, the exemplary illustrations may be modified according to manufacturing techniques and / or tolerances.

[0050] Unless otherwise defined, all terms (including technical and scientific terms) used in this disclosure have the same meanings as those commonly understood by those skilled in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and this disclosure, and will not be interpreted as having an idealized or overly formal meaning unless expressly defined in this disclosure.

[0051] Intelligent parking path planning aims to reduce the difficulty and time consumption of drivers during parking, and improve parking efficiency and safety. By integrating advanced information technology and algorithms, the intelligent parking system can perceive the surrounding environment in real time, including parking space location, obstacle distribution, traffic conditions, etc., and plan the optimal parking path accordingly to assist vehicles in quickly entering the parking space, saving parking time while ensuring the safety of drivers.

[0052] In related technologies, the algorithm overhead and resources of the on-board computing platform will affect the efficiency of successful path planning. Therefore, how to accurately plan parking paths and improve the parking success rate when available resources are limited has become a problem to be solved.

[0053] In view of this, embodiments of the present disclosure provide a parking path planning method and a parking path planning device, an electronic device, and a computer-readable medium.

[0054] In the disclosed embodiment, when planning a parking path, at least one first alternative parking path is determined based on the first path cost. Furthermore, the first alternative parking path can be optimized using the second path cost to obtain a better parking path, thereby improving the parking success rate.

[0055] In a first aspect, an embodiment of the present disclosure provides a parking path planning method.

[0056] Figure 1 This is a flow chart of a parking path planning method provided by an embodiment of the present disclosure. Figure 1 , the parking path planning method may include the following steps.

[0057] Step S101 : generating a parking grid map based on vehicle information of the vehicle to be parked and environmental information of the environment in which the vehicle to be parked is located. The parking grid map is used to indicate obstacle information and parking space information within a preset range.

[0058] Step S102 : Based on the parking grid map, at least one first alternative parking path is determined according to the first path cost. The first alternative parking path is a path starting from the vehicle to be parked and ending at an empty parking space.

[0059] Step S103 : For any first alternative parking path, optimize at least a portion of the first alternative parking path according to the second path cost to obtain a second alternative parking path corresponding to the first alternative parking path.

[0060] Step S104 : selecting a target parking path from the second candidate parking paths corresponding to the first candidate parking paths.

[0061] In some optional embodiments, the vehicle information of the vehicle to be parked includes position and posture information of the vehicle to be parked, wherein the position and posture information of the vehicle to be parked can represent the position and posture of the vehicle to be parked.

[0062] In some optional embodiments, a parking grid map is generated based on vehicle information of the vehicle to be parked and environmental information of the environment in which the vehicle to be parked is located, including: identifying target objects based on the environmental information of the environment in which the vehicle to be parked is located to obtain target object identification information, where the target objects include obstacles and vacant parking spaces; and generating a parking grid map based on preset map parameters and the posture information of the vehicle to be parked and the target object identification information.

[0063] As can be seen, first, obstacles and vacant parking spaces can be identified based on the environmental information of the environment in which the vehicle to be parked is located. Obstacles can be used to identify which locations are inaccessible, and vacant parking spaces can be used to identify which locations can accommodate the vehicle to be parked. On this basis, the location and posture of the vehicle to be parked are combined, and a parking grid map can be generated according to preset map parameters. The preset map parameters may include grid resolution, which can represent the size of a grid corresponding to the actual scene. For example, for the same target object, the number of grids occupied by the target object in the parking grid map will be different for different grid resolutions, and the corresponding path planning accuracy will also be different. In actual applications, the appropriate grid resolution can be selected according to needs, and the embodiments of the present disclosure are not limited to this.

[0064] In addition, the preset map parameters may further include a map size, which is used to represent the size of the parking grid map to be generated.

[0065] It should be noted that the above preset map parameters are only examples, and the embodiments of the present disclosure do not limit this.

[0066] In some optional embodiments, the environmental information includes multiple environmental images; accordingly, target object recognition is performed based on the environmental information of the environment in which the vehicle to be parked is located to obtain target object recognition information, including: inputting the multiple environmental images into a preset target detection model to obtain target object recognition information; wherein the target detection model is used to mark a target box in the environmental image, and the target box is used to select the target object in the environmental image.

[0067] In some optional embodiments, the preset target detection model is a YOLOv5 (You Only Look Once version 5) model, which is a model for implementing a target detection algorithm. The camera and sensor of the vehicle to be parked capture an environmental image, and the environmental image is input into the target detection model for processing to obtain a detection frame of the target object. Subsequently, the corresponding target object can be mapped to the parking grid map according to the coordinates of the detection frame.

[0068] For example, the target detection model is first loaded: the device information of the vehicle to be parked (such as CPU or GPU) is obtained, and the pre-trained YOLOv5 model is loaded, and information such as the model's stride and category names are obtained for backup. The image is then preprocessed: the input image file (such as the environment image) is read, the image is resized to the size required for model input (such as 640x640), and the aspect ratio of the image is maintained. The image is converted from HWC (height, width, channel) format to CHW (channel, height, width) format, and the image data is converted into a continuous array and converted into a PyTorch tensor. Furthermore, the preprocessed environment image is input into the YOLOv5 model for inference. During the processing, non-maximum suppression (NMS) can be used to process the model output, thereby removing redundant detection boxes and retaining the optimal detection results.

[0069] Based on the detection results, the coordinates of the detection boxes are converted from the scale of the model output back to the scale of the original image. The detection results are then traversed to extract the coordinates, confidence, and category information of each detection box. The detection boxes are drawn on the environment image and the category and confidence information are annotated. Finally, the environment image with the relevant information annotated is output.

[0070] In some optional embodiments, the activation function that can be used by the target detection model includes a ReLU (Rectified Linear Unit) function, thereby increasing the nonlinear expression ability of the model while reducing the problem of gradient disappearance.

[0071] In some optional embodiments, the loss function in the target detection model can represent positioning loss, classification loss, and confidence loss (Objectness Loss). Among them, the positioning loss can use the Complete Intersection over Union Loss (CIoU) function, which comprehensively considers the intersection over union ratio, aspect ratio, and center point distance between the predicted box and the true box, thereby more accurately measuring the difference between the predicted box and the true box.

[0072] Exemplarily, the CIoU loss function can be represented as: CIoULoss=1-IoU+\frac{d2}{c2}+alpha×v.

[0073] Among them, IoU is the intersection over union (IoU) between the predicted box and the true box, and its calculation formula is IoU = \frac{|A\capB|}{|A\cupB|}, where A and B are the areas of the predicted box and the true box respectively; d represents the Euclidean distance between the center points of the predicted box and the true box; c represents the diagonal length of the minimum closed box covering the predicted box and the true box; alpha is the weight coefficient used to balance the importance of different loss terms; v is the adjustment parameter used to adjust the loss function, which can be understood as a correction factor for the shape and direction of the target box, which is related to factors such as the aspect ratio of the predicted box and the true box; \frac represents the score function.

[0074] Classification loss (ClassLoss) can be expressed using the cross entropy loss function (CrossEntropyLoss) or focal loss (Focal Loss), which is used to characterize the difference between the predicted category and the true category. Focal Loss is a variant of the cross entropy loss function that can more effectively handle category imbalance problems.

[0075] Confidence loss can also use a cross-entropy loss function or focal loss to calculate the network's confidence, primarily used to determine whether a target object exists within the prediction frame. In some optional embodiments, the target object recognition information includes obstacle distribution information and position information of available parking spaces. Accordingly, based on preset map parameters, a parking grid map is generated based on the position information of the parked vehicle and the target object recognition information. This includes: generating an initial grid map based on the preset map parameters; and mapping the parked vehicle, obstacles, and available parking spaces onto the initial grid map based on the position information of the parked vehicle, obstacle distribution information, and position information of the available parking spaces to obtain the parking grid map. The obstacle distribution information may include the size, absolute position information, and / or relative position information of the obstacle. The absolute position information may be represented using the absolute position coordinates of the obstacle, while the relative position information may be represented based on, for example, the distance between the obstacle and the parked vehicle. The position information of the available parking spaces includes the position and posture of the available parking spaces.

[0076] For example, a blank initial grid map comprising multiple grids can be generated based on preset map parameters. Furthermore, based on the position and posture of the vehicle to be parked, the size and position of obstacles, and the position and posture of available parking spaces, the vehicles to be parked, obstacles, and available parking spaces can be mapped onto the initial grid map, thereby generating a parking grid map. Within this parking grid map, each grid can be assigned an attribute value, which can be used to clearly identify the grid as corresponding to the vehicle to be parked, obstacle, or available parking space.

[0077] For example, the grid attribute values include "0" and "1." "0" corresponds to a grid representing an optional path (including the vehicle to be parked and an open parking space), and "1" corresponds to a grid representing an obstacle. Later, when determining the parking path, the parking grid map can be combined with the location coordinates of the vehicle to be parked and the open parking space.

[0078] For example, the grid attribute values include "0," "1," "2," and "3." "0" represents a vehicle waiting to be parked, "1" represents an obstacle, "2" represents a drivable path, and "3" represents an empty parking space. Accordingly, the attribute value of each grid in the initial grid map is 0. Subsequently, by mapping the vehicles waiting to be parked, obstacles, and empty parking spaces to this initial grid map, the grid attribute values can be updated. The attribute value of the grid corresponding to the vehicle waiting to be parked in the map is updated to "0," the attribute value of the grid corresponding to the obstacle is updated to "1," the attribute value of the grid corresponding to the empty parking space is updated to "3," and the attribute value of the remaining grids corresponding to the drivable path is updated to "2." Furthermore, the parking path to be planned should start with the grid corresponding to the attribute value "0" and end with the grid corresponding to the attribute value "3," with the two connected by multiple grids with the attribute value "2." Since multiple attributes of the grids in the parking grid map can represent not only obstacles and drivable paths, but also the positions of vehicles to be parked and vacant parking spaces, a parking path can be searched directly based on the parking grid map.

[0079] In some optional embodiments, the first path cost includes the actual cost and the heuristic cost of the corresponding path, and the second path cost includes the actual cost of the corresponding path.

[0080] For example, the actual cost refers to the actual cost of the path from the starting point to the current node, which is mainly used to record the actual cost from the starting point to the current node to ensure the correctness of the path; the heuristic cost refers to the estimated value of the path cost from the current node to the target node, which is mainly used to estimate the cost from the current node to the target node, guide the search direction, and improve the efficiency of the path search.

[0081] For example, the actual cost may be determined according to the length of the path from the starting point to the current node.

[0082] For example, the actual cost can be determined based on the path lengths of multiple path segments from the starting point to the current node and the corresponding path segment weight coefficients, wherein the path segment weight coefficients can represent the driving difficulty, road condition complexity, etc. of the corresponding path segment.

[0083] Exemplarily, the heuristic cost may be determined based on the straight-line distance between the current node and the target node.

[0084] In some optional embodiments, determining at least one first alternative parking path based on a parking grid map and a first path cost includes: for an i-th processing step, determining, in the parking grid map, a plurality of extension nodes corresponding to an i-th parent node according to a preset extension parameter, where i is an integer greater than or equal to 1, and the first parent node is a mapping node of the vehicle to be parked in the parking grid map; determining a first path cost corresponding to each extension node; selecting an updated extension node from the plurality of extension nodes based on the first path cost corresponding to each extension node; if the updated extension node is not a target node, executing an i+1-th processing step using the updated extension node as an i+1-th parent node, where the target node is a mapping node of a vacant parking space in the parking grid map; and if the updated extension node is the target node, determining a first alternative parking path based on the first to i-th parent nodes and the updated extension node.

[0085] In some optional embodiments, the preset expansion parameter includes at least one of a preset expansion step length and a preset expansion angle, wherein the preset expansion step length represents the distance from the current node to the corresponding expansion node, and the preset expansion angle represents the angle from the current node to the corresponding expansion node.

[0086] For example, in a first processing step, the node corresponding to the vehicle to be parked in the parking grid map is used as the first parent node. Multiple extension nodes corresponding to the first parent node are determined according to preset extension parameters. A first path cost corresponding to each extension node is then determined, and the one with the smallest first path cost is selected from the multiple extension nodes as the updated extension node. A determination is then made as to whether the updated extension node belongs to the target node. If so, a first alternative parking path is determined based on the first parent node and the current updated extension node. If not, the updated extension node is used as the second parent node, and a second processing step is performed based on the second parent node. This process is repeated until an updated extension node is determined to belong to the target node. The first to last parent nodes and the last determined updated extension node are then concatenated to obtain the corresponding first alternative parking path.

[0087] For example, when determining the first path cost corresponding to a certain extended node, the actual cost can be determined based on the distance between the starting node and the extended node, and the heuristic cost can be determined based on the straight-line distance between the extended node and the target node. The actual cost and the heuristic cost are added together to obtain the first path cost corresponding to the extended node.

[0088] It should be noted that for each vacant parking space, at least one first alternative parking path corresponding to the vacant parking space can be determined. When there are multiple vacant parking spaces, multiple corresponding first alternative parking paths can be obtained.

[0089] After obtaining the first alternative parking path, in order to further improve path rationality, reduce path length, and increase parking success rate, the first alternative parking path can be optimized to obtain an optimized parking path (i.e., the second alternative parking path).

[0090] In some optional embodiments, optimizing at least a portion of the first alternative parking path based on the second path cost to obtain a second alternative parking path corresponding to the first alternative parking path includes: selecting a portion of the first alternative parking path from the first alternative parking path to obtain a first subpath during any optimization process of the first alternative parking path; determining a second subpath based on the second path cost, using the first node of the first subpath as a new starting point and the last node of the first subpath as a new end point; and replacing the first subpath of the first alternative parking path with the second subpath if the length of the second subpath is less than the length of the first subpath.

[0091] In other words, if there is a second sub-route that is better than the first sub-route, the second sub-route is used to replace the original first sub-route, thereby obtaining a better second alternative parking route.

[0092] For example, the first alternative parking path is ACFGHKTW, with nodes A, C, F, G, H, K, T, and W. If FGHK is selected as the first subpath, a path from F to K is searched based on the second path cost, using F as the new starting point and K as the new end point. This results in a second subpath FIK, which has a shorter length than the first. Therefore, the first subpath in the first alternative parking path is replaced with the second subpath, resulting in the second alternative parking path ACFIKTW. Comparing the first and second alternative parking paths, the path length is shorter, thus achieving path optimization.

[0093] It should be noted that for any first alternative parking path, the process of selecting and optimizing the first sub-path can be performed once or multiple times, and this is not limited in the present embodiment. During the j-th optimization (j>1), the path obtained by performing the j-1th optimization on the first alternative parking path is used as the optimized first alternative parking path, and the j+1th optimization is performed on this optimized first alternative parking path. For example, using the aforementioned first alternative parking path ACFGHKTW as an example, if multiple optimizations are performed on it, after the first optimization, the optimized first alternative parking path ACFIKTW can be obtained. During the second optimization, this optimized first alternative parking path ACFIKTW is used as the basis for optimization.

[0094] In some optional embodiments, the first candidate parking path may be determined based on an A* algorithm, and the first candidate parking path may be optimized based on a Dijkstra algorithm, thereby achieving efficient and reliable parking path planning.

[0095] For example, the current position of the vehicle to be parked and the position of the available parking space are obtained. For example, the position corresponding to the vehicle to be parked is (0,0), which represents the coordinates of the vehicle to be parked on the grid map and corresponds to the starting point of the path; the position corresponding to the available parking space is (4,4), which represents the coordinates of the available parking space where the vehicle to be parked needs to be parked and corresponds to the end point of the path.

[0096] Based on the above position information and obstacle information, a parking grid map can be obtained. The map can use a two-dimensional array, for example:

[0097] [0,2,2,2,1;

[0098] 2,1,1,2,1;

[0099] 2,2,2,2,1;

[0100] 2,1,1,3,2;

[0101] 2,2,2,2,2]

[0102] Among them, 0 represents the grid corresponding to the vehicle waiting to be parked; 1 represents the grid with obstacles and the vehicle cannot pass through; 2 represents the grid corresponding to the drivable area; and 3 represents the grid corresponding to the vacant parking space.

[0103] Define a node class, Node. Each node represents a grid in the parking grid map. The node class includes the node's position, the first path cost, and the corresponding parent node. For example, the node position is represented by the coordinates (x, y), the actual cost from the starting point to the current node is g, the actual cost from the starting point to the current node is the heuristic cost h, the first path cost f = g + h, and the parent node is represented by parent. In the subsequent path search process, the grid corresponding to "0" is the starting node, and the grid corresponding to "3" is the end node or target node.

[0104] Furthermore, a heuristic cost function is defined to calculate the heuristic cost. For example, the Euclidean distance formula can be used to calculate the straight-line distance from the current node to the target node as the heuristic cost h.

[0105] Based on this, the cost calculation is performed: the first open list is initialized, the starting point is added to the first open list, its g is set to 0, and h and f are calculated. The first path cost is determined based on h and f. The node with the lowest first path cost is taken from the first open list as the current node. The current node is determined to be the target node. If so, the path is backtracked and returned. If not, the four adjacent nodes of the current node (upper, lower, left, and right) are checked. For each adjacent node, if it is an empty grid and has not been visited, it is determined as an expansion node, and its g, h, and f are calculated. Its parent node is set as the current node and added to the first open list. The above process is repeated until the target node is found or the first open list is empty.

[0106] After the first alternative parking path is determined in the above manner, it can be optimized.

[0107] For example, the first subpath is selected from the first alternative parking path, and the new starting and ending points are determined. Based on this, the second open list is initialized, the starting point is added to the second open list, and its g is set to 0. The node with the lowest cost is taken from the second open list as the current node. A determination is made as to whether the current node is the target node. If so, the path is backtracked and returned. If not, the four adjacent nodes (upper, lower, left, and right) of the current node are checked. For each adjacent node, if it is an idle grid, the actual cost g from the starting point to the node is calculated. If the new g value is less than the g value of the current node, the g value of the node is updated, its parent node is set as the current node, and it is added to the second open list. This process is repeated until the target node is found or the second open list is empty.

[0108] After the above processing, a second alternative parking path can be determined. The path can be represented by a list containing node coordinates, such as [(0,0), (1,0), (2,0), (2,1), (2,2), (3,2), (4,2), (4,3), (4,4)], which represents the parking path of the vehicle to be parked from the starting point to the end point.

[0109] It should be noted that the A* algorithm can accelerate the search process through heuristic functions, reduce unnecessary node expansion, and improve path planning efficiency, making it particularly suitable for fast searches. However, it may not be globally optimal. The Dijkstra algorithm can ensure the shortest path in complex environments, especially when the A* algorithm is inaccurate. The Dijkstra algorithm can ensure the global optimality of the path. Based on this, combining the two to jointly determine the parking path can improve efficiency and accuracy on the one hand, combining the fast search capability of the A* algorithm with the global optimization capability of the Dijkstra algorithm.

[0110] In some optional embodiments, selecting a target parking path from among the second alternative parking paths corresponding to each first alternative parking path includes selecting a shortest path from among the plurality of second alternative parking paths as the target parking path, thereby parking the vehicle to be parked in an available parking space corresponding to the target parking path based on the target parking path.

[0111] For example, a plurality of second candidate parking paths correspond to a plurality of vacant parking spaces, and a second candidate parking path with the shortest path length is selected as the target parking path.

[0112] Figure 2 A flow chart of a parking path planning method provided by an embodiment of the present disclosure. Figure 2 , the parking path planning method may include the following steps.

[0113] Step S201 : performing target object recognition based on environmental information of the environment in which the vehicle to be parked is located to obtain target object recognition information.

[0114] Step S202: Generate an initial grid map based on preset map parameters.

[0115] In step S203, based on the position information of the parked vehicle, the obstacle distribution information, and the position information of the available parking spaces, the parked vehicle, obstacles, and available parking spaces are mapped into the initial grid map to obtain a parking grid map. In step S204, for the i-th processing step, multiple expansion nodes corresponding to the i-th parent node are determined in the parking grid map according to preset expansion parameters.

[0116] Wherein, i is an integer greater than or equal to 1, and the first parent node is the mapping node of the vehicle to be parked in the parking grid map.

[0117] Step S205: Determine the first path cost corresponding to each extended node.

[0118] Step S206 : selecting an updated extension node from the plurality of extension nodes according to the first path cost corresponding to each extension node.

[0119] Step S207: When the updated extended node does not belong to the target node, the updated extended node is taken as the (i+1)th parent node to perform the (i+1)th processing.

[0120] The target node is the mapping node of the vacant parking space in the parking grid map.

[0121] Step S208 : When the updated extension node belongs to the target node, a first alternative parking path is determined according to the 1st to ith parent nodes and the updated extension node.

[0122] Step S209 : For any optimization process of the first candidate parking path, a partial path is selected from the first candidate parking path to obtain a first sub-path.

[0123] Step S210 , taking the first node of the first subpath as a new starting point and the last node of the first subpath as a new end point, and determining a second subpath according to the second path cost.

[0124] Step S211 : When the path length of the second sub-path is less than the path length of the first sub-path, the first sub-path in the first candidate parking path is replaced by the second sub-path.

[0125] In step S212, upon completion of at least one optimization process, the current first alternative parking path is determined as the second alternative parking path. In step S213, the shortest path length among the plurality of second alternative parking paths is selected as the target parking path, and the vehicle to be parked is parked in an available parking space corresponding to the target parking path based on the target parking path.

[0126] The steps of the various methods above are divided only for the purpose of clear description. During implementation, they can be combined into one step or some steps can be split and decomposed into multiple steps. As long as they include the same logical relationship, they are all within the scope of protection of this patent. Adding insignificant modifications or introducing insignificant designs to the algorithm or process without changing the core design of the algorithm and process are all within the scope of protection of this patent.

[0127] In a second aspect, an embodiment of the present disclosure provides a parking path planning path.

[0128] Figure 3 A block diagram of a parking path planning device provided by an embodiment of the present disclosure. Figure 3 The parking path planning device 300 may include the following modules.

[0129] A generating module 301 is configured to generate a parking grid map based on vehicle information of the vehicle to be parked and environmental information of the environment in which the vehicle to be parked is located. The parking grid map is configured to indicate obstacle information and parking space information within a preset range.

[0130] A determination module 302 is configured to determine, based on the parking grid map and according to the first path cost, at least one first alternative parking path, the first alternative parking path being a path starting from the vehicle to be parked and ending at an empty parking space;

[0131] an optimization module 303 for optimizing, for any first alternative parking path, at least a portion of the first alternative parking path according to the second path cost, to obtain a second alternative parking path corresponding to the first alternative parking path;

[0132] The selection module 304 is configured to select a target parking path from the second candidate parking paths corresponding to the first candidate parking paths.

[0133] In an embodiment of the present disclosure, a generation module generates a parking grid map based on vehicle information of a vehicle to be parked and environmental information of the environment in which the vehicle is parked. The parking grid map indicates obstacle information and parking space information within a preset range. A determination module determines, based on the parking grid map, at least one first alternative parking path according to a first path cost. The first alternative parking path is a path starting from the vehicle to be parked and ending at an available parking space. An optimization module optimizes at least a portion of the first alternative parking path according to a second path cost for each first alternative parking path to obtain a second alternative parking path corresponding to the first alternative parking path. A selection module selects a target parking path from the second alternative parking paths corresponding to each first alternative parking path. Thus, the present embodiment can determine at least one first alternative parking path based on the first path cost and optimize the first alternative parking path using the second path cost, thereby improving the parking success rate. The functions or modules included in the device provided in the embodiments of the present disclosure can be used to execute the method described in the first aspect of the method embodiment above. Its specific implementation and technical effects can be referred to the description of the above method embodiment. For the sake of brevity, they will not be repeated here.

[0134] It should be noted that all modules involved in this embodiment are logical modules. In actual applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. In addition, to highlight the innovations of this disclosure, this embodiment does not include units that are not closely related to solving the technical problems proposed by this disclosure. However, this does not mean that other units do not exist in this embodiment.

[0135] Figure 4 A block diagram of an electronic device provided in an embodiment of the present disclosure.

[0136] Reference Figure 4 , an embodiment of the present disclosure provides an electronic device, comprising:

[0137] One or more processors 901;

[0138] a memory 402 storing one or more programs, which, when executed by one or more processors, enable the one or more processors to implement any one of the above-mentioned parking path planning methods;

[0139] One or more I / O interfaces 403 are connected between the processor and the memory and are configured to implement information exchange between the processor and the memory.

[0140] Among them, the processor 401 is a device with data processing capabilities, including but not limited to a central processing unit (CPU); the memory 402 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically SDRAM, DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), and flash memory (FLASH); the I / O interface (read-write interface) 403 is connected between the processor 401 and the memory 402, and can realize information interaction between the processor 401 and the memory 402, including but not limited to a data bus (Bus), etc.

[0141] In some embodiments, the processor 401 , the memory 402 , and the I / O interface 403 are connected to each other via a bus, and further connected to other components of the computing device.

[0142] This embodiment further provides a computer-readable medium having a computer program stored thereon. When the program is executed by a processor, the parking path planning method provided in this embodiment is implemented. To avoid repeated description, the specific steps of the parking path planning method are not repeated here.

[0143] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods invented above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In hardware implementations, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As is well known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable, and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those skilled in the art that communication media typically embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0144] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0145] Those skilled in the art will understand that although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features from different embodiments is meant to be within the scope of the present embodiment and to form different embodiments.

[0146] It is understood that the above embodiments are merely exemplary embodiments for illustrating the principles of the present disclosure, and the present disclosure is not limited thereto. Those skilled in the art may make various modifications and improvements without departing from the spirit and substance of the present disclosure, and such modifications and improvements are also considered to be within the scope of protection of the present disclosure.

Claims

1. A parking path planning method, characterized in that: include: generating a parking grid map based on vehicle information of the vehicle to be parked and environmental information of the environment in which the vehicle to be parked is located, wherein the parking grid map is used to indicate obstacle information and parking space information within a preset range; determining, based on the parking grid map and according to a first path cost, at least one first alternative parking path, the first alternative parking path being a path starting from the vehicle to be parked and ending at an empty parking space; For any of the first alternative parking paths, optimizing at least a portion of the first alternative parking paths according to a second path cost to obtain a second alternative parking path corresponding to the first alternative parking path; A target parking path is selected from the second candidate parking paths corresponding to each of the first candidate parking paths.

2. The method according to claim 1, characterized in that The vehicle information of the vehicle to be parked includes the position information of the vehicle to be parked; Generating a parking grid map based on vehicle information of the vehicle to be parked and environmental information of an environment in which the vehicle to be parked is located includes: Performing target object recognition based on environmental information of the environment in which the vehicle to be parked is located to obtain target object recognition information, wherein the target objects include obstacles and vacant parking spaces; The parking grid map is generated based on preset map parameters and according to the position information of the vehicle to be parked and the target object recognition information.

3. The method according to claim 2, characterized in that The environmental information includes a plurality of environmental images; The step of identifying a target object based on environmental information of an environment in which the vehicle to be parked is located to obtain target object identification information includes: Inputting the plurality of environment images into a preset target detection model to obtain target object recognition information; The target detection model is used to mark a target frame in the environment image, and the target frame is used to select a target object in the environment image.

4. The method according to claim 2, characterized in that The target object recognition information includes the distribution information of obstacles and the position information of the vacant parking space; The generating of the parking grid map based on preset map parameters and according to the position information of the vehicle to be parked and the target object recognition information includes: Based on the preset map parameters, generating an initial grid map; According to the posture information of the vehicle to be parked, the obstacle distribution information and the posture information of the vacant parking spaces, the vehicle to be parked, the obstacles and the vacant parking spaces are mapped into the initial grid map to obtain the parking grid map.

5. The method according to claim 1, wherein The first path cost includes the actual cost and the heuristic cost of the corresponding path, and the second path cost includes the actual cost of the corresponding path.

6. The method according to claim 1, characterized in that The determining, based on the parking grid map and according to a first path cost, at least one first candidate parking path includes: For the i-th processing step, a plurality of expansion nodes corresponding to the i-th parent node are determined in the parking grid map according to a preset expansion parameter, where i is an integer greater than or equal to 1, and the first parent node is a mapping node of the vehicle to be parked in the parking grid map; Determine the first path cost corresponding to each extended node; selecting an updated extension node from the plurality of extension nodes according to the first path cost corresponding to each of the extension nodes; In the case where the updated extended node does not belong to the target node, performing the (i+1)th processing process with the updated extended node as the (i+1)th parent node, the target node being a mapping node of the vacant parking space in the parking grid map; In a case where the updated extended node belongs to the target node, the first candidate parking path is determined according to the 1st to i-th parent nodes and the updated extended node.

7. The method according to claim 1 or 6, characterized in that The optimizing at least a portion of the first candidate parking path according to the second path cost to obtain a second candidate parking path corresponding to the first candidate parking path includes: In any optimization process of the first candidate parking path, a portion of the first candidate parking path is selected to obtain a first sub-path; Taking the first node of the first subpath as a new starting point and the last node of the first subpath as a new end point, determining a second subpath according to the second path cost; When the path length of the second sub-path is shorter than the path length of the first sub-path, the first sub-path in the first candidate parking path is replaced by the second sub-path.

8. A parking path planning device, characterized in that: include: a generating module for generating a parking grid map based on vehicle information of the vehicle to be parked and environmental information of the environment in which the vehicle to be parked is located, wherein the parking grid map is used to indicate obstacle information and parking space information within a preset range; a determining module configured to determine, based on the parking grid map and according to a first path cost, at least one first alternative parking path, the first alternative parking path being a path starting from the vehicle to be parked and ending at an empty parking space; an optimization module configured to optimize, for any of the first alternative parking paths, at least a portion of the first alternative parking paths according to a second path cost, to obtain a second alternative parking path corresponding to the first alternative parking path; A selection module is configured to select a target parking path from the second candidate parking paths corresponding to each of the first candidate parking paths.

9. An electronic device, characterized in that: include: one or more processors; a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the parking path planning method according to any one of claims 1 to 7; One or more I / O interfaces are connected between the processor and the memory and configured to implement information interaction between the processor and the memory.

10. A computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the parking path planning method according to any one of claims 1 to 7 is implemented.

Images