An Automated Parking Path Planning Method Based on Comprehensive Driving Cost

Abstract

This application discloses an automated parking path planning method based on comprehensive driving costs. It acquires parking environment information and, combined with initial vehicle position information, constructs a state graph model that satisfies vehicle kinematic constraints. Based on this, it determines the comprehensive driving cost by integrating factors such as vehicle travel distance, steering changes, obstacle influences, and direction of movement switching. Then, it performs path search on the state graph model based on the comprehensive driving cost to generate an automated parking trajectory. This application combines vehicle kinematic constraints with multi-dimensional driving costs, making the planned automated parking trajectory more consistent with the actual driving characteristics of the vehicle. This effectively reduces frequent steering and gear shifting operations, significantly improving the comfort and safety of automated parking, and increasing the safety margin between the vehicle and obstacles, thereby improving the safety, smoothness, and path planning quality during the automated parking process.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to an automatic parking path planning method based on comprehensive driving costs. Background Technology

[0002] With the development of intelligent driving technology, automated parking systems (APS) have been widely applied in intelligent vehicles. Path planning is a crucial step in achieving automated parking, and its results directly affect whether the vehicle can safely and smoothly enter the target parking space. Currently, automated parking path planning typically employs graph search algorithms such as Dijkstra's, with path length as the primary optimization objective. However, in actual parking processes, considering only path length is insufficient to simultaneously account for various driving factors such as vehicle kinematic constraints, steering angle changes, obstacle safety distances, and forward / reverse switching. This can easily lead to excessive steering, frequent gear shifts, or the vehicle getting too close to obstacles, thus affecting the safety, comfort, and path quality of the parking process. Summary of the Invention

[0003] The purpose of this application is to overcome the above-mentioned problems and provide an automatic parking path planning method, an automatic parking system, a storage medium, an electronic device, and a computer program product based on comprehensive driving costs.

[0004] The technical solution of this application provides an automatic parking path planning method based on comprehensive driving costs, including: Obtain parking environment information and initial vehicle location information. The parking environment information includes the target parking space information. Construct a state graph model that satisfies vehicle kinematic constraints based on parking environment information and initial vehicle position information; Determine the comprehensive driving cost based on the state diagram model; The automatic parking trajectory is obtained by performing path search on the state graph model based on the comprehensive driving cost.

[0005] Furthermore, a state graph model satisfying vehicle kinematic constraints is constructed based on parking environment information and initial vehicle position information, including: Multiple state nodes are determined based on the target parking space information and the initial vehicle position information; Motion edges connecting each state node are generated using motion primitives; The state graph model is obtained based on the state nodes and motion edges.

[0006] Furthermore, the state nodes include the starting state node, the target state node, and the intermediate state nodes; Multiple state nodes are identified, including: The initial state node and the target state node are determined based on the initial vehicle location information and the target parking space information; Multiple intermediate state nodes are generated based on the starting state node and the target state node.

[0007] Furthermore, the comprehensive driving cost is determined based on the state diagram model, including: Obtain the basic driving cost, steering cost, obstacle cost, and direction switching cost corresponding to each motion edge in the state graph model; The comprehensive driving cost corresponding to each movement edge is determined based on the basic driving cost, steering cost, obstacle cost, and movement direction switching cost.

[0008] Furthermore, path search is performed on the state graph model based on the comprehensive driving cost, including: Initialize path search; The node search and update are performed according to the comprehensive driving value cycle until the preset termination condition is met. Automatic parking trajectories are generated based on node update results.

[0009] Further, the path search is initialized, including: Add the starting state node to the priority queue; Initialize the cumulative path cost corresponding to the starting state node to zero.

[0010] Furthermore, node search and node updates are performed based on the comprehensive driving value, including: The state node with the minimum cumulative path cost is selected from the priority queue and designated as the current state node. Iterate through the adjacent state nodes corresponding to the current state node and update the cumulative path cost; Repeat the above steps until the preset termination condition is met.

[0011] Furthermore, update the cumulative path cost, including: Calculate the cumulative path cost of candidate paths from the current state node to adjacent state nodes; In response to a candidate cumulative path cost being less than the cumulative path cost of an adjacent state node, update the cumulative path cost of the adjacent state node.

[0012] Furthermore, after updating the cumulative path cost of the adjacent state node, the process also includes: Add the adjacent state node to the priority queue and record the current state node as the predecessor node of the adjacent state node.

[0013] Further, the cumulative path cost of candidate paths from the current state node to adjacent state nodes is calculated, including: The candidate cumulative path cost is obtained by adding the cumulative path cost of the current state node to the comprehensive driving cost corresponding to the motion edge connecting the current state node and the adjacent state nodes.

[0014] Furthermore, the preset termination conditions include: The target state node has been found, or the priority queue is empty.

[0015] The technical solution of this application also provides an automatic parking system based on comprehensive driving costs, including: The acquisition module is used to acquire parking environment information; The modeling module is used to construct a state graph model that satisfies vehicle kinematic constraints based on parking environment information and initial vehicle position information. The cost calculation module is used to determine the overall driving cost based on the state diagram model; The path planning module is used to perform path search on the state graph model based on the comprehensive driving cost to obtain the automatic parking trajectory.

[0016] The technical solution of this application also provides a storage medium that stores computer instructions. When the computer executes the computer instructions, it is used to execute the aforementioned automatic parking path planning method based on comprehensive driving costs.

[0017] The technical solution of this application also provides an electronic device, including at least one processor; and, A memory that is communicatively connected to at least one processor; wherein, The memory stores instructions that can be executed by at least one processor, which enables the at least one processor to perform an automated parking path planning method based on comprehensive driving costs, as described above.

[0018] The technical solution of this application also provides a computer program product, including a computer program / instruction, characterized in that, when the computer program / instruction is executed by a processor, it implements an automatic parking path planning method based on comprehensive driving costs as described in the previous one.

[0019] The above technical solution has the following beneficial effects: This application presents an automated parking path planning method based on comprehensive driving costs. It acquires parking environment information and, combined with initial vehicle position information, constructs a state graph model that satisfies vehicle kinematic constraints. Based on this, it determines the comprehensive driving cost by integrating factors such as vehicle travel distance, steering changes, obstacle influences, and direction of movement switching. Then, it performs path search on the state graph model based on the comprehensive driving cost to generate an automated parking trajectory. This application combines vehicle kinematic constraints with multi-dimensional driving costs, making the planned automated parking trajectory more consistent with the actual driving characteristics of the vehicle. This effectively reduces frequent steering and gear shifting operations, significantly improving the comfort and safety of automated parking, and increasing the safety margin between the vehicle and obstacles, thereby enhancing the safety, smoothness, and path planning quality during the automated parking process. Attached Figure Description

[0020] The disclosure of this application will become more readily understood with reference to the accompanying drawings. It should be understood that these drawings are for illustrative purposes only and are not intended to limit the scope of protection of this application. In the drawings: Figure 1 This is a flowchart of an automatic parking path planning method based on comprehensive driving costs in one embodiment of this application; Figure 2 This is a flowchart of an automatic parking path planning method based on comprehensive driving costs in one embodiment of this application; Figure 3 This is a schematic diagram of an automatic parking system based on comprehensive driving costs in one embodiment of this application; Figure 4 This is a schematic diagram of the hardware structure of an electronic device in one embodiment of this application. Detailed Implementation

[0021] The specific embodiments of this application will be further described below with reference to the accompanying drawings.

[0022] It is readily understood that, based on the technical solution of this application, various structural and implementation methods can be interchanged by those skilled in the art without altering the essential spirit of this application. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this application and should not be considered as the entirety of this application or as limitations or restrictions on the technical solution of the application.

[0023] The directional terms such as up, down, left, right, front, back, front, back, top, and bottom mentioned or possibly used in this specification are defined relative to the structures shown in the accompanying drawings. These are relative concepts and may therefore vary depending on their location and usage. Therefore, these or other directional terms should not be interpreted as restrictive. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0024] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meanings of the above in this application according to the specific circumstances.

[0025] like Figure 1 The flowchart shown below illustrates an automatic parking path planning method based on comprehensive driving costs in one embodiment of this application, specifically including: Step S101: Obtain parking environment information and initial vehicle location information. The parking environment information includes target parking space information. Step S102: Construct a state graph model that satisfies vehicle kinematic constraints based on parking environment information and initial vehicle position information; Step S103: Determine the comprehensive driving cost based on the state diagram model; Step S104: Perform path search on the state graph model based on the comprehensive driving cost to obtain the automatic parking track.

[0026] Specifically, in step S101, environmental information surrounding the vehicle can be acquired through onboard sensors, which may include one or more of ultrasonic radar, surround-view cameras, millimeter-wave radar, or lidar. A parking environment map is constructed based on the acquired environmental information, which includes at least the parking environment map, target parking space information, obstacle information, and vehicle-accessible area information. Simultaneously, the vehicle's current initial position information is acquired to provide environmental and initial state information for subsequent path planning.

[0027] In step S102, a state graph model is constructed based on the parking environment information and the initial vehicle position information. Specifically, this involves determining the initial vehicle state node and the target state node corresponding to the target parking space, and generating multiple intermediate state nodes within the vehicle's passable area by combining vehicle kinematic constraints such as wheelbase and minimum turning radius. Motion primitives that satisfy the vehicle kinematic constraints are used to connect the state nodes, forming the state graph model. These motion primitives can be generated using Reeds-Shepp curves, Dubins curves, or discrete control quantity integration to ensure that the motion trajectories between state nodes meet the actual driving constraints of the vehicle.

[0028] In step S103, the comprehensive driving cost value is determined based on the state graph model. This is achieved by determining the basic driving cost, steering cost, obstacle cost, and direction-switching cost based on the motion characteristics corresponding to the motion edges between each state node in the state graph model, and then fusing them according to preset weights to obtain the comprehensive driving cost value. Specifically, the basic driving cost characterizes the distance the vehicle travels along the motion edge, the steering cost characterizes the degree of change in the vehicle's steering angle, the obstacle cost is determined based on the distance between the vehicle and the obstacle, and the direction-switching cost characterizes the operational cost incurred when switching between forward and reverse directions. This ensures that the comprehensive driving cost value reflects the actual driving cost of the vehicle traversing the corresponding motion edge, providing an evaluation basis for subsequent path searching.

[0029] In step S104, a path search is performed on the state graph model based on the comprehensive driving cost value to obtain the automatic parking trajectory. Specifically, the initial state node is used as the search starting point, and node search and updates are performed iteratively based on the comprehensive driving cost value to gradually determine the cumulative path cost corresponding to each state node. When the preset search termination condition is met, the automatic parking trajectory is generated based on the node update results. The comprehensive driving cost value serves as the search cost for each motion edge in the state graph model, ensuring that the path search process comprehensively considers factors such as vehicle travel distance, steering changes, obstacle influences, and motion direction switching, thereby obtaining an automatic parking trajectory that satisfies vehicle kinematic constraints.

[0030] This application presents an automated parking path planning method based on comprehensive driving costs. It acquires parking environment information and, combined with initial vehicle position information, constructs a state graph model that satisfies vehicle kinematic constraints. Based on this, it determines the comprehensive driving cost by integrating factors such as vehicle travel distance, steering changes, obstacle influences, and direction switching. Then, it performs path search on the state graph model based on the comprehensive driving cost to generate an automated parking trajectory. This application combines vehicle kinematic constraints with multi-dimensional driving costs, making the planned automated parking trajectory more consistent with the actual driving characteristics of the vehicle. This effectively reduces frequent steering and gear shifting operations, significantly improving the comfort and safety of automated parking, and increasing the safety margin between the vehicle and obstacles, thereby enhancing the safety, smoothness, and path planning quality during the automated parking process.

[0031] In one embodiment, a state graph model satisfying vehicle kinematic constraints is constructed based on parking environment information and initial vehicle position information, including: Multiple state nodes are determined based on the target parking space information and the initial vehicle position information; Motion edges connecting each state node are generated using motion primitives; The state graph model is obtained based on the state nodes and motion edges.

[0032] In this embodiment, state nodes represent a reachable state of the vehicle during path planning. Each state node contains not only the vehicle's location but also its orientation information, thus accurately describing the vehicle's actual posture in the parking environment. Motion edges connect two adjacent state nodes, representing feasible movement processes by which the vehicle can move from one state node to another. The state graph model includes multiple state nodes and motion edges connecting them, describing all feasible movement routes of the vehicle in the parking environment.

[0033] In this embodiment, a state graph model satisfying vehicle kinematic constraints is constructed based on parking environment information and initial vehicle position information. First, multiple state nodes are determined based on the target parking space information and initial vehicle position information. Motion primitives are used to generate motion edges connecting these state nodes, establishing a corresponding motion relationship between any two directly reachable state nodes. Since the motion edges are generated under the premise of satisfying vehicle kinematic constraints, each connected state node can be transformed through the vehicle's actual executable motion methods. Finally, a state graph model is constructed based on the state nodes and motion edges, converting feasible motion relationships in the parking environment into a graph structure representation, providing a data foundation for subsequent path search.

[0034] This embodiment constructs a state graph model using state nodes and motion edges, ensuring that all motion edges in the state graph model meet the actual driving constraints of the vehicle. This avoids generating motion trajectories that the vehicle cannot execute during subsequent path search, thereby improving the feasibility of path planning.

[0035] In one embodiment, the state node includes a starting state node, a target state node, and intermediate state nodes; Multiple state nodes are identified, including: The initial state node and the target state node are determined based on the initial vehicle location information and the target parking space information; Multiple intermediate state nodes are generated based on the starting state node and the target state node.

[0036] In this embodiment, the initial state node refers to the initial state position of the vehicle when it begins automatic parking, determined by the vehicle's current state and orientation. The target state node refers to the target state of the vehicle after automatic parking is completed, determined by the position of the target parking space and the vehicle's orientation after parking. Intermediate state nodes refer to the multiple intermediate states that the vehicle may pass through during its movement from the initial state node to the target state node. Each intermediate state node contains the vehicle's position and orientation and satisfies vehicle kinematic constraints.

[0037] In this embodiment, the initial state node and the target state node are first determined based on the initial vehicle position information and the target parking space information, respectively. Specifically, the initial position and orientation of the vehicle are determined based on the initial vehicle position information and used as the initial state node; the target pose of the vehicle after parking is determined based on the target parking space information and used as the target state node. Subsequently, based on the initial state node and the target state node, multiple intermediate state nodes are sampled and generated within the free space accessible to the vehicle.

[0038] This embodiment generates multiple intermediate state nodes, discretizing the continuous movement process of the vehicle from the starting position to the target parking space into multiple analyzable vehicle states. This can completely describe the key states in the automatic parking process, providing a foundation for the subsequent construction of a state diagram model, thereby improving the rationality and feasibility of path planning.

[0039] In another embodiment, determining the comprehensive driving cost based on the state diagram model includes: Obtain the basic driving cost, steering cost, obstacle cost, and direction switching cost corresponding to each motion edge in the state graph model; The comprehensive driving cost corresponding to each movement edge is determined based on the basic driving cost, steering cost, obstacle cost, and movement direction switching cost.

[0040] In this embodiment, the basic driving cost characterizes the fundamental cost required for the vehicle to travel along the moving edge, primarily reflecting the distance the vehicle travels along that edge. The steering cost characterizes the degree of change in steering angle during the vehicle's journey along the moving edge; a greater change in steering angle results in a higher steering cost. The obstacle cost characterizes the safe distance between the vehicle and obstacles during the journey along the moving edge; the closer the vehicle is to the obstacle, the higher the obstacle cost. The direction-switching cost characterizes the operational cost incurred when the vehicle shifts gears forward or backward during movement. The comprehensive driving cost reflects the actual driving cost of the vehicle traversing the corresponding moving edge, and is determined jointly by the basic driving cost, steering cost, obstacle cost, and direction-switching cost.

[0041] In this embodiment, the comprehensive driving cost is determined based on a state diagram model. Specifically, the basic driving cost, steering cost, obstacle cost, and direction-switching cost corresponding to each motion edge in the state diagram model are first obtained. The basic driving cost is determined based on the arc length of the motion edge, the steering cost is determined based on the change in steering angle, the obstacle cost is determined based on the distance between the vehicle and the nearest obstacle during its travel along the motion edge, and the direction-switching cost is determined based on whether the vehicle switches between forward and backward. Subsequently, these costs are weighted and fused according to preset weights to obtain the comprehensive driving cost corresponding to each motion edge, ensuring that each motion edge has an evaluation result that comprehensively reflects the actual driving cost of the vehicle.

[0042] This embodiment comprehensively considers various factors such as vehicle travel distance, steering changes, obstacle effects, and movement direction switching to more fully reflect the overall driving cost during actual vehicle operation. This makes the cost of each moving edge more consistent with actual driving needs, thereby improving the safety, smoothness, and feasibility of subsequent path planning results.

[0043] In one specific embodiment, the basic driving cost is the arc length C_distance corresponding to the moving edge; the steering cost C_steering is proportional to the angle change formed by the vehicle's orientation changing from the current state node to the adjacent state node, so as to increase the cost of large angles and frequent steering; the obstacle cost C_obstacle is determined based on the distance between the vehicle and the nearest obstacle during the vehicle's travel along the moving edge. When the obstacle distance is smaller, the corresponding cost increases in a non-linear manner to improve path safety; the motion direction switching cost C_direction is determined based on whether the vehicle switches between forward and backward. When a motion direction switch occurs, a fixed penalty is added to reduce unnecessary gear shifting operations.

[0044] The formula for calculating the overall driving cost is as follows: ; Where Cost(e) is the comprehensive driving cost corresponding to the moving edge, C_distance is the basic driving cost, C_steering is the steering cost, C_obstacle is the obstacle cost, C_direction is the driving direction switching cost, and a, b, c, and d represent the weight coefficients corresponding to the basic driving cost, steering cost, obstacle cost, and driving direction switching cost, respectively. Each weight coefficient can be configured according to the actual automatic parking needs to adjust the degree of influence of different driving costs in the comprehensive driving cost.

[0045] In one specific embodiment, the arc length L corresponding to a certain moving edge is 3 m, the vehicle orientation change angle Δθ is 30°, the minimum distance between the vehicle and the nearest obstacle during the vehicle's travel along the moving edge is 0.5 m, and the current movement direction is backward while the previous movement direction of the moving edge is forward, thus a movement direction switch occurs.

[0046] Let the weighting coefficients be: base driving cost weight a = 1.0, steering cost weight b = 0.5, and obstacle cost defined as c = 2.0 × (1 / distance). 2 The cost of switching the direction of motion is d=3.0.

[0047] The formula for the overall driving cost value is as follows: .

[0048] In another embodiment, path search is performed on the state graph model based on the comprehensive driving cost, including: Initialize path search; The node search and update are performed according to the comprehensive driving value cycle until the preset termination condition is met. Automatic parking trajectories are generated based on node update results.

[0049] In this embodiment, node search refers to selecting the current state node to be expanded in the state graph model and determining its neighboring state nodes. Node update refers to updating the cumulative path cost corresponding to neighboring state nodes based on the comprehensive driving cost and recording the relationship between state nodes.

[0050] In this embodiment, path search is performed on the state graph model based on the comprehensive driving cost value. Specifically, the path search process is first initialized to determine the starting state. Then, node search and node updates are performed cyclically based on the comprehensive driving cost value. In each cycle, the current state node to be expanded is selected, and the cumulative path cost of adjacent state nodes is updated according to the comprehensive driving cost value, gradually expanding the path search towards the direction with the lowest comprehensive driving cost. When a preset termination condition is met, the path search process ends, and an automatic parking trajectory is generated based on the state node relationships recorded during the node update process.

[0051] This embodiment employs a cyclical node search and update approach, enabling the vehicle to continuously compare the comprehensive driving costs corresponding to different motion paths throughout the entire state graph model. It gradually selects paths with lower comprehensive driving costs, rather than relying solely on local information for path selection. This approach takes into account various driving factors such as driving distance, steering changes, obstacle influences, and changes in motion direction, thereby improving the rationality, safety, and feasibility of the automatic parking trajectory.

[0052] In one embodiment, initializing the path search includes: Add the starting state node to the priority queue; Initialize the cumulative path cost corresponding to the starting state node to zero.

[0053] In this embodiment, the priority queue is used to store the state nodes to be searched. The state nodes are stored according to their priority, and the search order is determined according to the magnitude of the cumulative path cost. The cumulative path cost refers to the cumulative comprehensive driving cost of the path from the starting state node to the current state node, which is used to characterize the comprehensive driving cost required for the vehicle to reach the current state node.

[0054] In this embodiment, the path search process is initialized. Specifically, the starting state node is first added to a priority queue so that the path search can use the vehicle's current state as the starting point. Then, the cumulative path cost corresponding to the starting state node is initialized to zero, indicating that the vehicle has not yet traversed any moving edges when it is at the starting state node, and therefore there is no cumulative driving cost. After initialization, subsequent node searches and node updates are performed.

[0055] This embodiment uses the starting state node as the initial node for path search and initializes its cumulative path cost to zero. This ensures that the cumulative path cost of all subsequent state nodes is gradually accumulated based on the starting state node, avoiding deviations in path search results due to inconsistent initial path cost settings, thereby improving the accuracy and consistency of path search results.

[0056] In one embodiment, node search and node update are performed based on the comprehensive driving value, including: The state node with the minimum cumulative path cost is selected from the priority queue and designated as the current state node. Iterate through the adjacent state nodes corresponding to the current state node and update the cumulative path cost; Repeat the above steps until the preset termination condition is met.

[0057] In this embodiment, the current state node refers to the state node that has been taken from the priority queue and is currently undergoing an expanded search, and whose cumulative path cost is the lowest in the priority queue. Adjacent state nodes refer to state nodes that can be directly connected to the current state node via motion edges, and can be considered as candidate nodes for path searching. In this embodiment, node search and node update are performed iteratively based on the comprehensive driving cost value. Specifically, firstly, the state node with the minimum cumulative path cost is selected from the priority queue as the current state node, and the search for paths with lower comprehensive driving costs is expanded. Then, each adjacent state node corresponding to the current state node is traversed, and the cumulative path cost of the adjacent state nodes is updated based on the comprehensive driving cost value corresponding to each motion edge. After completing the search for the current state node, the process of selecting the state node with the minimum cumulative path cost from the priority queue is repeated until a preset termination condition is met.

[0058] This embodiment prioritizes searching for state nodes with lower cumulative path costs and continuously updates the cumulative path costs of adjacent state nodes. This allows the path search process to continuously expand towards a lower overall driving cost, gradually approaching the optimal path without traversing all possible paths, thus improving path search efficiency.

[0059] In one embodiment, updating the cumulative path cost includes: Calculate the cumulative path cost of candidate paths from the current state node to adjacent state nodes; In response to a candidate cumulative path cost being less than the cumulative path cost of an adjacent state node, update the cumulative path cost of the adjacent state node.

[0060] In this embodiment, the candidate cumulative path cost refers to the cumulative path cost when a vehicle reaches an adjacent state node from the current state node. It is the comprehensive driving cost when the current search path reaches that adjacent node.

[0061] In this embodiment, the cumulative path cost of adjacent state nodes is updated. First, the candidate cumulative path cost from the current state node to the adjacent state node is calculated, and then compared with the cumulative path cost currently stored in the adjacent state node. When the candidate cumulative path cost is less than the cumulative path cost of the adjacent state node, it means that a path with a lower overall driving cost can be obtained by traveling from the current state node to the adjacent state node, so the candidate cumulative path cost is used to replace the original cumulative path cost; otherwise, the original cumulative path cost remains unchanged.

[0062] Specifically, when a vehicle searches for a current state node, it can reach the same adjacent state node through different paths. First, the cumulative cost of candidate paths from the current state node to the adjacent state node is calculated and compared with the currently stored cumulative path cost of that adjacent state node. If the candidate cumulative path cost is lower, it indicates that the current search path has a lower overall driving cost, so the cumulative path cost corresponding to the adjacent state node is updated; if the candidate cumulative path cost is higher, the original cumulative path cost remains unchanged. Because the system always retains paths with lower overall driving costs, it can avoid poor paths affecting subsequent search results, thus improving the overall quality of automatic parking path planning.

[0063] This embodiment compares the candidate cumulative path cost with the existing cumulative path cost, retaining only the path with the lower overall driving cost. This avoids paths with higher overall driving costs from participating in subsequent searches, ensuring that the cumulative path cost always remains among the best results of the currently searched paths. This improves the accuracy of path search and provides a reliable foundation for generating automatic parking trajectories with better overall driving performance.

[0064] In one embodiment, after updating the cumulative path cost of the adjacent state node, the method further includes: Add the adjacent state node to the priority queue and record the current state node as the predecessor node of the adjacent state node.

[0065] In this embodiment, the predecessor node refers to the previous state node before reaching the current state node during the path search process, and is used to record the relationship between state nodes.

[0066] In this embodiment, after the cumulative path cost of an adjacent state node is updated, the adjacent state node is added to a priority queue, and the current state node is recorded as the predecessor node of the adjacent state node. Specifically, adding an adjacent state node to the priority queue allows it to participate in path search as a subsequent node to be searched; at the same time, by recording the connection relationship between the current state node and the adjacent state node, path association information between state nodes is established, providing a basis for the subsequent generation of automatic parking trajectories.

[0067] Specifically, as subsequent path searches continue, the system can sequentially use a priority queue to search each state node and gradually reconstruct the complete path based on the recorded predecessor node relationships. Because the connection relationships between state nodes are continuously preserved during the search process, the system can ultimately generate an accurate automatic parking trajectory from the starting state node to the target state node, improving the accuracy and completeness of the path planning results.

[0068] This embodiment adds the updated adjacent state nodes back to the priority queue, allowing state nodes with better overall driving costs to continue participating in subsequent path searches. Recording the predecessor node can completely preserve the connection relationships between each state node during the path search process, avoiding the inability to record the complete path later. This ensures that the automatic parking trajectory can accurately reflect the path search results and improves the reliability of automatic parking path planning.

[0069] In another embodiment, calculating the candidate cumulative path cost from the current state node to the adjacent state node includes: The candidate cumulative path cost is obtained by adding the cumulative path cost of the current state node to the comprehensive driving cost corresponding to the motion edge connecting the current state node and the adjacent state nodes.

[0070] In this embodiment, the motion edge refers to the feasible motion path between the current state node and the adjacent state node, which is used to represent the motion process of the vehicle from the current state node to the adjacent state node. Each motion edge corresponds to a comprehensive driving cost value.

[0071] In this embodiment, the candidate cumulative path cost from the current state node to the adjacent state node is calculated. Specifically, the cumulative path cost corresponding to the current state node is added to the comprehensive driving cost corresponding to the motion edge connecting the current state node and the adjacent state node to obtain the candidate cumulative path cost from the current state node to the adjacent state node. Here, the cumulative path cost corresponding to the current state node represents the comprehensive driving cost already incurred by the vehicle from the starting state node to the current state node, and the comprehensive driving cost corresponding to the motion edge represents the additional comprehensive driving cost of the vehicle continuing to move from the current state node to the adjacent state node. The sum of the two yields the total cumulative path cost of the vehicle reaching the adjacent state node along the current path.

[0072] This embodiment obtains the candidate cumulative path cost for the vehicle to reach adjacent state nodes by summing the cumulative path cost corresponding to the current state node with the comprehensive driving cost corresponding to the moving edge. This allows the driving cost generated by the current moving edge to be included in the overall path cost, making the path evaluation more consistent with the actual driving process of the vehicle, and improving the rationality and accuracy of path planning.

[0073] In another embodiment, the preset termination condition includes: The target state node has been found, or the priority queue is empty.

[0074] In this embodiment, the preset termination conditions include the target state node being found or the priority queue being empty. During the path search process, state nodes are continuously selected from the priority queue for searching. When the target state node is found, it means that a feasible path from the starting state node to the target state node has been obtained, and the path search ends at this time. If the priority queue is empty, it means that there are no more state nodes that can be searched, that is, no feasible path to the target state node has been found, and the path search also ends.

[0075] Specifically, during the automatic parking process, path searching begins from the initial state node and continuously traverses each state node in the state graph model. When the target state node is reached during the search, it indicates that an automatic parking path that allows the vehicle to enter the target parking space has been found, and the path search immediately ends. If an obstacle in the parking environment completely blocks the vehicle from entering the target parking space, as the path search continues, the number of state nodes waiting to be searched in the priority queue gradually decreases until the priority queue is empty. At this point, it is determined that there is no automatic parking path that meets the requirements, and the current path search ends. This avoids invalid searches and improves the operational efficiency and reliability of automatic parking path planning.

[0076] This embodiment sets two termination conditions. On the one hand, when the target state node is found, the search can be terminated in time to avoid traversing irrelevant state nodes and improve the efficiency of path planning. On the other hand, when the priority queue is empty, it can be determined in time that there is no path that meets the requirements, to avoid the program from continuing to perform invalid searches and improve the integrity and reliability of the path search process.

[0077] In one embodiment, before obtaining the automatic parking trajectory, the method further includes: The target path obtained from the path search is smoothed to obtain the automatic parking trajectory.

[0078] In this embodiment, the target path obtained through path search is smoothed to obtain an automatic parking trajectory. This is because the target path obtained through path search consists of multiple state nodes and corresponding motion edges, and there may be significant changes in direction between adjacent path segments. Smoothing the target path allows for a smoother transition between adjacent path segments, thereby forming a continuous automatic parking trajectory.

[0079] This embodiment smooths the target path, reducing situations where there are large local directional changes in the path, making the automatic parking trajectory more continuous and smooth, more in line with the actual driving characteristics of the vehicle, reducing sudden steering changes during parking, and improving the feasibility and driving comfort of the automatic parking trajectory.

[0080] like Figure 2 As shown, a flowchart of an automatic parking path planning method based on comprehensive driving cost in one embodiment of this application is presented, specifically including: Step S201: Obtain parking environment information and initial vehicle location information. The parking environment information includes target parking space information. Step S202: Determine the starting state node and the target state node based on the initial vehicle position information and the target parking space information; Step S203: Generate multiple intermediate state nodes based on the starting state node and the target state node; Step S204: Generate motion edges connecting each state node using motion primitives; Step S205: Obtain the state graph model based on the state nodes and motion edges; Step S206: Obtain the basic driving cost, steering cost, obstacle cost, and movement direction switching cost corresponding to each motion edge in the state graph model; Step S207: Determine the comprehensive driving cost corresponding to each movement edge based on the basic driving cost, steering cost, obstacle cost, and movement direction switching cost; Step S208: Add the starting state node to the priority queue; Step S209: Initialize the cumulative path cost corresponding to the starting state node to zero; Step S210: Take the state node with the minimum cumulative path cost from the priority queue as the current state node; Step S211: Traverse the adjacent state nodes corresponding to the current state node and update the cumulative path cost; Step S212: Calculate the candidate cumulative path cost from the current state node to the adjacent state node; Step S213: In response to the candidate cumulative path cost being less than the cumulative path cost of the adjacent state node, update the cumulative path cost of the adjacent state node; Step S214: Add the adjacent state node to the priority queue and record the current state node as the predecessor node of the adjacent state node; Step S215: Repeat the above steps until the preset termination condition is met to obtain the automatic parking trajectory.

[0081] like Figure 3As shown, embodiments of this application also provide an automated parking system based on comprehensive driving costs, including: Module 301 is used to acquire parking environment information; Modeling module 302 is used to construct a state diagram model that satisfies vehicle kinematic constraints based on parking environment information and initial vehicle position information; Cost calculation module 303 is used to determine the comprehensive driving cost based on the state diagram model; The path planning module 304 is used to perform path search on the state graph model based on the comprehensive driving cost to obtain the automatic parking trajectory.

[0082] An automatic parking system based on comprehensive driving costs in this embodiment includes an acquisition module 301, a modeling module 302, a cost calculation module 303, and a path planning module 304.

[0083] The acquisition module 301 is used to acquire parking environment information and initial vehicle location information. The parking environment information includes at least a parking environment map, target parking space information, obstacle information, and vehicle passable area information, providing environmental data for subsequent route planning.

[0084] The modeling module 302 is used to construct a state graph model that satisfies vehicle kinematic constraints based on parking environment information and initial vehicle position information. Specifically, the modeling module 302 determines the initial state node and the target state node based on the initial vehicle position information and the target parking space information, and generates multiple intermediate state nodes within the vehicle's passable area; it uses motion primitives that satisfy vehicle kinematic constraints to generate motion edges connecting each state node, forming a state graph model, where each motion edge stores corresponding distance and motion information such as changes in vehicle orientation.

[0085] The cost calculation module 303 is used to determine the comprehensive driving cost value based on the state diagram model. Specifically, the cost calculation module 303 obtains the basic driving cost, steering cost, obstacle cost, and movement direction switching cost corresponding to each movement edge, and performs weighted calculation according to preset weights to obtain the comprehensive driving cost value corresponding to each movement edge.

[0086] The path planning module 304 is used to perform path search on the state graph model based on the comprehensive driving cost value to obtain the automatic parking trajectory. Specifically, the path planning module 304 initializes the path search, adds the starting state node to the priority queue, and uses the starting state node as the search starting point. It then iteratively performs node search and node update based on the comprehensive driving cost value corresponding to each motion edge until the target state node is found or the priority queue is empty. Finally, it generates the automatic parking trajectory based on the node update results.

[0087] Embodiments of this application also provide a storage medium storing computer instructions, which, when executed by a computer, are used to perform an automatic parking path planning method based on comprehensive driving costs as described in any of the preceding embodiments.

[0088] Figure 4 An electronic device according to this application is shown, comprising: At least one processor 401; and, Memory 402 is communicatively connected to at least one processor 401; wherein, The memory 402 stores instructions that can be executed by at least one processor 401, which enables the at least one processor 401 to perform all steps of an automated parking path planning method based on comprehensive driving costs in any of the foregoing method embodiments.

[0089] Figure 4 Taking a processor 401 as an example: The electronic device may also include an input device 403 and an output device 404.

[0090] The processor 401, memory 402, input device 403 and output device 404 can be connected by a bus or other means. The figure shows an example of connection by bus.

[0091] Memory 402, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to an automatic parking path planning method based on comprehensive driving costs in an embodiment of this application. Figure 1 and Figure 2 The method flow is shown. The processor 401 executes various functional applications and data processing by running non-volatile software programs, instructions, and modules stored in the memory 402, thereby implementing an automatic parking path planning method based on comprehensive driving costs in the above embodiments.

[0092] Memory 402 may include a program storage area and a data storage area. The program storage area may store an operating system and applications required for at least one function. The data storage area may store data created based on the use of an automated parking path planning method based on comprehensive driving costs. Furthermore, memory 402 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, memory 402 may optionally include memory remotely located relative to processor 401, which can be connected via a network to a device performing an automated parking path planning method based on comprehensive driving costs. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0093] Input device 403 can receive user clicks and generate signal inputs related to user settings and function control for authentication methods based on Ethernet end-to-end communication. Output device 404 may include display devices such as a display screen.

[0094] One or more modules are stored in memory 402, and when run by one or more processors 401, they execute an automatic parking path planning method based on comprehensive driving costs in any of the above method embodiments.

[0095] Embodiments of this application also provide a computer program product, including a computer program / instructions that, when executed by a processor, implement an automatic parking path planning method based on comprehensive driving costs as described in any of the preceding embodiments.

[0096] As needed, the above technical solutions can be combined to achieve the best technical effect.

[0097] The above are merely the principles and preferred embodiments of this application. It should be noted that, for those skilled in the art, implementation methods obtained by appropriately combining the technical solutions disclosed in different embodiments are also included within the technical scope of this invention. Based on the principles of this application, several other modifications can also be made, which should also be considered within the protection scope of this application.

Claims

1. An automatic parking path planning method based on comprehensive driving cost, characterized in that, include: Obtain parking environment information and initial vehicle location information. The parking environment information includes the target parking space information. Construct a state graph model that satisfies vehicle kinematic constraints based on parking environment information and initial vehicle position information; Determine the comprehensive driving cost based on the state diagram model; The automatic parking trajectory is obtained by performing path search on the state graph model based on the comprehensive driving cost.

2. The method according to claim 1, characterized in that, Based on parking environment information and initial vehicle position information, a state graph model satisfying vehicle kinematic constraints is constructed, including: Multiple state nodes are determined based on the target parking space information and the initial vehicle position information; Motion edges connecting each state node are generated using motion primitives; The state graph model is obtained based on the state nodes and motion edges.

3. The method according to claim 2, characterized in that, The state nodes include the starting state node, the target state node, and the intermediate state nodes; Multiple state nodes are identified, including: The initial state node and the target state node are determined based on the initial vehicle location information and the target parking space information; Multiple intermediate state nodes are generated based on the starting state node and the target state node.

4. The method according to claim 1, characterized in that, The overall driving cost is determined based on the state diagram model, including: Obtain the basic driving cost, steering cost, obstacle cost, and direction switching cost corresponding to each motion edge in the state graph model; The comprehensive driving cost corresponding to each movement edge is determined based on the basic driving cost, steering cost, obstacle cost, and movement direction switching cost.

5. A method according to claim 1, characterized in that, Path search is performed on the state graph model based on the comprehensive driving cost, including: Initialize path search; The node search and update are performed according to the comprehensive driving value cycle until the preset termination condition is met. Automatic parking trajectories are generated based on node update results.

6. A method according to claim 5, characterized in that, Initialize path search, including: Add the starting state node to the priority queue; Initialize the cumulative path cost corresponding to the starting state node to zero.

7. A method according to claim 5, characterized in that, Node search and update are performed based on the comprehensive driving value, including: The state node with the minimum cumulative path cost is selected from the priority queue and designated as the current state node. Iterate through the adjacent state nodes corresponding to the current state node and update the cumulative path cost; Repeat the above steps until the preset termination condition is met.

8. A method according to claim 7, characterized in that, Update cumulative path cost, including: Calculate the cumulative path cost of candidate paths from the current state node to adjacent state nodes; In response to a candidate cumulative path cost being less than the cumulative path cost of an adjacent state node, update the cumulative path cost of the adjacent state node.

9. A method according to claim 8, characterized in that, After updating the cumulative path cost of the adjacent state node, the following steps are also included: Add the adjacent state node to the priority queue and record the current state node as the predecessor node of the adjacent state node.

10. A method according to claim 8, characterized in that, Calculate the candidate cumulative path cost from the current state node to its adjacent state nodes, including: The candidate cumulative path cost is obtained by adding the cumulative path cost of the current state node to the comprehensive driving cost corresponding to the motion edge connecting the current state node and the adjacent state nodes.

11. A method according to claim 5, characterized in that, The preset termination conditions include: The target state node has been found, or the priority queue is empty.

12. An automated parking system based on comprehensive driving costs, characterized in that, include: The acquisition module is used to acquire parking environment information; The modeling module is used to construct a state graph model that satisfies vehicle kinematic constraints based on parking environment information and initial vehicle position information. The cost calculation module is used to determine the overall driving cost based on the state diagram model; The path planning module is used to perform path search on the state graph model based on the comprehensive driving cost to obtain the automatic parking trajectory.

13. A storage medium, characterized in that, The storage medium stores computer instructions, which, when executed by the computer, are used to perform an automatic parking path planning method based on comprehensive driving costs as described in any one of claims 1-11.

14. An electronic device, characterized in that, Includes at least one processor; and, A memory that is communicatively connected to at least one processor; wherein, The memory stores instructions that can be executed by at least one processor to enable the at least one processor to perform an automatic parking path planning method based on comprehensive driving costs as claimed in any one of claims 1-11.

15. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement an automatic parking path planning method based on comprehensive driving costs as described in any one of claims 1-11.

Images