Automatic parking method and related equipment

Abstract

The invention discloses an automatic parking method and related equipment, relates to the field of intelligent driving and automotive electronics, and mainly aims to solve the problem that the existing automatic parking technology is not accurate enough in evaluation and planning of parking spaces. The method comprises the steps that under the condition that a parking task is triggered, the parking space selection cost of each selectable parking space of a target vehicle is calculated, and the parking space selection cost is used for representing the parking difficulty degree of the parking space relative to the target vehicle and the matching degree of the parking space and the user requirement; a candidate parking space list is selected based on the comparison result of the parking space selection cost of each parking space and a cost threshold value, and the cost threshold value is determined based on the speed of the target vehicle; and path pre-planning is carried out on each parking space in the candidate parking space list based on a preset path planning process, so that the planning cost of each parking space in the candidate parking space list can be acquired, the planning cost is used for evaluating the advantages and disadvantages of the path of parking the corresponding parking space from the current vehicle position, and the parking space with the lowest planning cost is the target parking space.

Description

Technical Field

[0001] This invention relates to the fields of intelligent driving and automotive electronics, and in particular to an automatic parking method and related equipment. Background Technology

[0002] Automated parking technology, as an important direction for the development of automotive intelligence, aims to provide users with convenient parking solutions. However, existing automated parking systems still have significant limitations in their decision-making and planning capabilities in practical applications. After triggering a parking task, current systems typically evaluate available parking spaces based solely on the single dimension of the straight-line distance between the vehicle and the space, simply ranking and recommending them. This overly simplistic evaluation mechanism cannot comprehensively and accurately reflect the true difficulty of parking in a space. Existing solutions lack an efficient and robust pre-planning mechanism to pre-assess the advantages and disadvantages of different parking paths. The entire system's decision-making process is highly arbitrary, ultimately resulting in low accuracy in overall decision-making and planning. This has become a key technical bottleneck restricting the improvement of user experience and functional performance. Summary of the Invention

[0003] In view of the above problems, the present invention provides an automatic parking method and related equipment, the main purpose of which is to solve the problem that the existing automatic parking technology is not accurate enough in assessing and planning parking spaces.

[0004] To solve at least one of the above-mentioned technical problems, in a first aspect, the present invention provides an automatic parking method, the method comprising: When a parking task is triggered, the parking space selection cost for each available parking space for the target vehicle is calculated, wherein the parking space selection cost is used to characterize the ease or difficulty of parking the parking space relative to the target vehicle and the degree of matching with the user's needs. A list of candidate parking spaces is selected based on a comparison between the parking space selection cost and a cost threshold for each parking space, wherein the cost threshold is determined based on the speed of the target vehicle. Based on a preset path planning process, a path pre-planning is performed for each parking space in the candidate parking space list to obtain the planning cost of each parking space in the candidate parking space list. The planning cost is used to evaluate the merits of the path from the current vehicle position to the corresponding parking space, and the parking space with the lowest planning cost is the target parking space.

[0005] Optionally, the factors influencing the cost of parking space selection include: Distance assessment results, which are used to calculate the straight-line distance from the center of the rear axle of the vehicle to the center of the parking space entrance; Angle evaluation results, which are used to calculate the angle deviation between the vehicle's current driving direction and the parking space entrance direction; Width assessment results, which are used to assess parking spaces whose actual width is less than the standard width; Environmental assessment results, which are used to calculate the obstacle density around the parking space; Personalized assessment results are used to adjust the parking space selection cost based on vehicle status and user habits.

[0006] Optionally, when a parking task is triggered, calculating the parking space selection cost for each available parking space for the target vehicle includes: Obtain the distance assessment result, angle assessment result, width assessment result, environmental assessment result, and personalized assessment result for the current calculated parking space; Based on the distance assessment results, the angle assessment results, the width assessment results, the environmental assessment results, and the personalized assessment results, the parking space selection cost for the current calculated parking space is calculated using a predefined cost function.

[0007] Optionally, the above methods also include: The cost threshold is calculated based on the cost threshold, the current speed, the speed limit, and the speed coupling coefficient. The cost threshold represents the maximum acceptable cost when the vehicle is stationary. The speed limit is the speed limit that triggers the parking task. The speed coupling coefficient is used to adjust the influence of speed on the threshold. The cost threshold is directly proportional to the vehicle speed.

[0008] Optionally, the selection of the candidate parking space list based on the comparison result of the parking space selection cost and the cost threshold for each parking space includes: Based on the comparison between the parking space selection cost and the cost threshold for each parking space, parking spaces with a selection cost less than or equal to the cost threshold are selected. The selected parking spaces are sorted from lowest to highest cost, and one or more parking spaces with the lowest cost are selected to form the candidate parking space list.

[0009] Optionally, the preset path planning adopts a hierarchical planning strategy, and the algorithm execution order of the hierarchical planning strategy is as follows: geometric splicing algorithm, hybrid A and RS curve algorithm, and hybrid A and RS segmentation optimization algorithm. The preset path planning process performs path pre-planning for each parking space in the candidate parking space list to obtain the planning cost for each parking space in the candidate parking space list, including: For each parking space in the candidate parking space list, the algorithm is called sequentially according to the algorithm execution order; If the algorithm planning for the current priority is successful, use its planning result and terminate the subsequent algorithm calls; If the algorithm planning of the current priority fails, the algorithm of the next priority is started to continue planning until the algorithm planning is successful; The planning cost of the currently calculated parking space is calculated based on the successful planning results of the algorithm.

[0010] Optionally, the above methods also include: When the target vehicle automatically parks in the target parking space, the deviation between the actual driving trajectory of the target vehicle and the pre-planned reference trajectory is monitored based on the Hausdorf distance. If the Hausdorf distance exceeds a preset Hausdorf distance threshold, safety measures are triggered, including deceleration and replanning.

[0011] Secondly, embodiments of the present invention also provide an automatic parking device, comprising: The calculation unit is used to calculate the parking space selection cost of each available parking space for the target vehicle when a parking task is triggered, wherein the parking space selection cost is used to characterize the parking difficulty of the parking space relative to the target vehicle and the degree of matching with the user's needs. A selection unit is used to select a list of candidate parking spaces based on a comparison between the parking space selection cost and a cost threshold for each parking space, wherein the cost threshold is determined based on the speed of the target vehicle. The planning unit is used to perform path pre-planning for each parking space in the candidate parking space list based on a preset path planning process, so as to obtain the planning cost of each parking space in the candidate parking space list. The planning cost is used to evaluate the merits of the path from the current vehicle position to the corresponding parking space, and the parking space with the lowest planning cost is the target parking space.

[0012] To achieve the above objectives, according to a third aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium comprising a stored program, wherein, when the program is executed by a processor, the steps of the above-described automatic parking method are implemented.

[0013] To achieve the above objectives, according to a fourth aspect of the present invention, an electronic device is provided, comprising at least one processor and at least one memory connected to the processor; wherein the processor is configured to invoke program instructions in the memory to execute the steps of the above-described automatic parking method.

[0014] By employing the above technical solutions, the automatic parking method and related equipment provided by this invention address the problem of insufficient accuracy in the evaluation and planning of parking spaces in existing automatic parking technologies. This invention calculates the parking space selection cost for each available parking space when a parking task is triggered, where the parking space selection cost characterizes the ease of parking the target vehicle relative to the target vehicle and its matching degree with user needs. A candidate parking space list is selected based on a comparison between the parking space selection cost of each parking space and a cost threshold, where the cost threshold is determined based on the speed of the target vehicle. A pre-planned path is performed for each parking space in the candidate parking space list based on a preset path planning process to obtain the planning cost for each parking space in the candidate parking space list. The planning cost is used to evaluate the merits of the path from the current vehicle position to the corresponding parking space, and the parking space with the lowest planning cost is the target parking space.

[0015] In the above scheme, a comprehensive parking space selection cost concept is first introduced to replace the traditional single-distance assessment. This cost, as a comprehensive indicator, is designed to characterize the overall parking difficulty and suitability of the parking space itself, thus generating a more comprehensive and accurate preliminary score for each parking space at the decision-making stage. Based on this, a dynamic cost threshold linked to vehicle speed is introduced for screening. This allows the system to have different screening tolerances under different driving conditions. Finally, the scheme initiates substantive path pre-planning for the limited number of candidate parking spaces selected, that is, simulating and calculating a feasible parking path for each candidate parking space and calculating a planning cost for horizontal comparison of the path's merits. By simulating the actual parking process, it quantifies the dynamic execution cost of the path and incorporates it into the final decision, thereby ensuring that the selected target parking space is not only theoretically easy to park in, but also practically reachable through an efficient and smooth path. This progressive decision-making chain, from "comprehensive preliminary assessment" to "dynamic coarse screening" and then to "path simulation verification," reduces the risk of decision bias caused by missing evaluation dimensions and a lack of forward-looking path verification.

[0016] Accordingly, the automatic parking device, equipment, and computer-readable storage medium provided in the embodiments of the present invention also have the above-mentioned technical effects.

[0017] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0018] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating an automatic parking method provided by an embodiment of the present invention is shown; Figure 2 This diagram illustrates the composition of an automatic parking device according to an embodiment of the present invention. Figure 3 This diagram illustrates the composition of an automatic parking electronic device according to an embodiment of the present invention. Detailed Implementation

[0019] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0020] To address the issue of insufficient accuracy in parking space assessment and planning in existing automated parking technologies, this invention provides an automated parking method, such as... Figure 1 As shown, the method includes: S101. When a parking task is triggered, calculate the parking space selection cost for each available parking space for the target vehicle, wherein the parking space selection cost is used to characterize the ease or difficulty of parking the parking space relative to the target vehicle and the degree of matching with the user's needs. For example, the parking space selection cost described above is used to characterize the ease of parking a parking space relative to the target vehicle and its match with the user's needs. It is achieved through a comprehensive evaluation by simultaneously considering information from multiple dimensions. By quantifying the ease of parking and the match with user needs into a comparable cost, the parking space selection cost allows for more accurate sorting and filtering of parking spaces in the early stages of decision-making. This reduces the risk of incorrectly recommending parking spaces due to incomplete information or biased evaluation, thereby improving the overall decision-making quality and user experience of the automatic parking system.

[0021] In one embodiment, the factors influencing the parking space selection cost include: Distance assessment results, which are used to calculate the straight-line distance from the center of the rear axle of the vehicle to the center of the parking space entrance; Angle evaluation results, which are used to calculate the angle deviation between the vehicle's current driving direction and the parking space entrance direction; Width assessment results, which are used to assess parking spaces whose actual width is less than the standard width; Environmental assessment results, which are used to calculate the obstacle density around the parking space; Personalized assessment results are used to adjust the parking space selection cost based on vehicle status and user habits.

[0022] For example, the distance assessment result refers to the straight-line distance from the center of the vehicle's rear axle to the center of the parking space entrance, used to quantify the distance between the parking space and the vehicle; the angle assessment result refers to the angular deviation between the vehicle's current driving direction and the direction of the parking space entrance, used to measure the difficulty of adjusting the vehicle's posture; the width assessment result evaluates the spatial adaptability of the parking space by comparing the actual width of the parking space with the standard width, and applies corresponding penalties when the actual width is insufficient; the environmental assessment result calculates the obstacle density around the parking space through sensor data to reflect the complexity of the parking environment; and the personalized assessment result dynamically adjusts the priority of the parking space based on the vehicle's real-time status, such as remaining battery power, and the user's historical habits, such as charging preferences, to make the recommended results more in line with actual needs.

[0023] When calculating the cost of parking space selection, this application simultaneously acquires the evaluation results of the above five dimensions and integrates them into a unified quantitative index through a predefined cost function. For example, in scenarios where the vehicle is an electric vehicle with low remaining battery power, the personalized evaluation result will automatically increase the priority of parking spaces with charging piles, thereby significantly reducing the overall selection cost of such parking spaces and ensuring that the system can prioritize recommending parking spaces that meet the user's urgent needs during the screening process. At the same time, for narrow parking spaces whose actual width is significantly smaller than the standard width, the width evaluation result will generate a higher penalty value, increasing the cost of the parking space and thus reducing its likelihood of being recommended, thereby automatically avoiding high-risk parking spaces.

[0024] The cost of parking space selection The calculation formula is as follows:

[0025] The Euclidean distance from the center of the vehicle's rear axle to the center of the parking space entrance represents the distance to the parking space. In this variable, α > 0. The weighting coefficient α determines the proportion of distance in the total cost. A larger value indicates a stronger impact of distance on the overall cost—that is, for every additional unit of Euclidean distance, the total cost increases significantly. In parking scenarios, the value of α needs to be determined based on the path characteristics of different parking scenarios (such as perpendicular parking spaces and parallel parking spaces). Figure 2 As shown, adjacent parking spaces for lateral parking are often farther apart, so α can be appropriately reduced to avoid excessively amplifying the distance cost; perpendicular parking spaces are closer together, so α can be increased to prioritize the shorter distance path.

[0026] : Deviation between vehicle heading angle and parking space entrance direction ( , For vehicle direction, In this variable, β>0 represents the weighting coefficient for angular deviation, determining its impact on cost. A larger value indicates that the algorithm "rejects" large angular deviations and prioritizes paths with a heading closer to the parking space. The physical significance in a parking scenario is that it measures the difficulty and cost of adjusting vehicle posture. The calibration of β needs to match the vehicle's minimum turning radius to avoid path redundancy caused by excessively pursuing small deviations.

[0027] Parking space width penalty item ( , (for standard width) Standard parking space width must match the vehicle model design (e.g., compact car). =2.5m, medium and large-sized vehicles =2.7m); The actual measured width of the current parking space is determined in real time by the fusion of LiDAR and visual sensors. Width penalty coefficient is an empirically calibrated parameter. k >0, used to adjust the non-linear growth rate of the penalty intensity. When ≥ hour, ≤0, result of exponential function ≤1, the penalty has almost no effect; the wider the parking space, the smaller the penalty; when < hour, >0, the exponential function will non-linearly amplify the penalty value (e.g., =0.8, When =0.3m, =e0.24≈1.27; when the difference expands to 0.5m, =e0.4≈1.49), to avoid recommending parking spaces that are too narrow, which could lead to the risk of scratches during parking. γ>0 determines the proportion of the width penalty in the total cost. The larger the value, the more sensitive the algorithm is to penalizing parking spaces whose actual width deviates from the standard width. During the parking space selection and sorting process, it will be more inclined to select parking spaces whose width meets or exceeds the standard requirements, in order to reduce the risk of scratches and posture adjustment costs caused by space constraints during parking.

[0028] The density of obstacles around the parking space is quantified using data fused from radar and visual sensors. = ,in To detect the number of obstacle points within the area, The total area of ​​the detection region is δ. δ>0 is the weighting coefficient for obstacle density, used to measure the impact of parking environment complexity on the overall cost. The larger the value of δ, the more significant the algorithm's tendency to avoid areas with dense obstacles, prioritizing parking spaces with open environments and high obstacle avoidance redundancy to improve the safety and stability of the parking process.

[0029] This is a multi-dimensional weighting coefficient that integrates vehicle status, parking space attributes, and user driving habits. It quantifies the match between parking spaces and user needs; the higher the value, the higher the priority of the parking space in the cost function, and the higher the corresponding cost term. The smaller the proportion, the better. The core design logic of this parameter is to achieve a parking space recommendation strategy that combines "scenario adaptation and personalized customization".

[0030] Taking parking spaces with charging stations as an example, the specific dimensions and association rules are as follows: Based on the vehicle's remaining battery power Establish a rule linking battery level threshold and parking space type as the core trigger condition: set a user's habitual charging battery level threshold. (This threshold is obtained by the vehicle system through learning from historical charging data, such as when the user frequently...) (Charging starts when the battery level is ≤20%) When the vehicle's remaining battery power is in real time ≤ At that time, the system automatically raises the charging station parking space. The weight value and the magnitude of the increase can be determined by the calibration coefficient. Regulation, i.e. = (1+ )( As the basic weight of parking spaces, >1); when At that time, the weight of the charging pile parking space returns to the basic value, and the selection criteria are consistent with those of ordinary parking spaces.

[0031] The core value of this mechanism lies in prioritizing users' choice of charging parking spaces when vehicle battery is low, avoiding range anxiety caused by depleted battery and improving the parking experience for new energy vehicles. This parameter can be further expanded to incorporate multiple reference factors, such as parking space type preference (based on users' historical parking data, different types of parking spaces are assigned differentiated weights, which can be flexibly expanded to various forms such as perpendicular parking spaces, parallel parking spaces, and angled parking spaces) and parking direction preference (for the two core parking directions of front-end parking and rear-end parking, direction preference weights are assigned based on user operating habits and parking space characteristics).

[0032] By employing the aforementioned technical solution and introducing a multi-factor fusion evaluation mechanism, the cost of parking space selection can simultaneously consider multiple characteristics such as geometric adaptation, environmental complexity, and personalized needs. This expands the system's evaluation of parking spaces from a single distance dimension to a more comprehensive dimension, thereby generating a more accurate and realistic score for each parking space at the decision-making stage. This effectively reduces the possibility of incorrect parking space recommendations due to incomplete or biased evaluation information, and improves the rationality and reliability of parking space selection.

[0033] In one embodiment, calculating the parking space selection cost for each available parking space for the target vehicle when a parking task is triggered includes: Obtain the distance assessment result, angle assessment result, width assessment result, environmental assessment result, and personalized assessment result for the current calculated parking space; Based on the distance assessment results, the angle assessment results, the width assessment results, the environmental assessment results, and the personalized assessment results, the parking space selection cost for the current calculated parking space is calculated using a predefined cost function.

[0034] For example, the distance assessment results, angle assessment results, width assessment results, environmental assessment results, and personalized assessment results mentioned above are input information for multiple dimensions on which the cost of parking space selection is based.

[0035] Upon triggering a parking task, this application synchronously acquires data such as distance assessment, angle assessment, width assessment, environmental assessment, and personalized assessment for each available parking space. These results are then passed as input to a predefined cost function, which performs a comprehensive calculation to determine the parking space selection cost. For example, when the vehicle is an electric vehicle with low remaining battery power, the personalized assessment automatically prioritizes parking spaces with charging stations, positively adjusting the personalized assessment results for such spaces and thus reducing their overall selection cost through the cost function. Simultaneously, for parking spaces with insufficient width, the width assessment generates a higher penalty value, resulting in a larger cost value output by the cost function, thus placing them lower in the ranking.

[0036] By integrating multi-source evaluation results and using a cost function for unified quantification, the system can generate a comprehensive index that fully reflects the difficulty of parking and the matching degree of user needs. This expands the parking space evaluation process from a single dimension to multi-factor fusion, improving the comprehensiveness and accuracy of the evaluation, effectively reducing the risk of decision-making errors caused by incomplete information or insufficient evaluation, and enhancing the reliability and user experience of the automatic parking system in the parking space selection stage.

[0037] S102. A candidate parking space list is selected based on the comparison result of the parking space selection cost and the cost threshold for each parking space, wherein the cost threshold is determined based on the speed of the target vehicle. In one embodiment, the above method further includes: The cost threshold is calculated based on the cost threshold, the current speed, the speed limit, and the speed coupling coefficient. The cost threshold represents the maximum acceptable cost when the vehicle is stationary. The speed limit is the speed limit that triggers the parking task. The speed coupling coefficient is used to adjust the influence of speed on the threshold. The cost threshold is directly proportional to the vehicle speed.

[0038] For example, the aforementioned cost threshold refers to the maximum acceptable cost when the vehicle is stationary. As a benchmark for threshold calculation, it can be dynamically adjusted according to the real-time speed of the vehicle and is directly proportional to the vehicle speed. The higher the vehicle speed, the higher the acceptable cost threshold. The current speed refers to the real-time speed of the target vehicle, which is dynamically obtained by the on-board system. The speed limit refers to the maximum speed limit allowed when the automatic parking function is triggered. The speed coupling coefficient is a calibration parameter used to adjust the degree of influence of speed changes on the cost threshold.

[0039] This application calculates the cost threshold by comprehensively considering the cost threshold, current speed, speed limit, and speed coupling coefficient. The cost threshold is directly proportional to the vehicle speed; that is, the higher the vehicle speed, the higher the calculated cost threshold, and vice versa. For example, when the vehicle is traveling at a high speed, the system automatically increases the cost threshold, allowing more parking spaces to meet the screening criteria and enter the candidate list, adapting to fast search scenarios. Conversely, when the vehicle decelerates or approaches a standstill, the cost threshold decreases, focusing on a few optimal parking spaces to ensure recommendation accuracy.

[0040] The cost threshold The calculation formula is as follows:

[0041] : Base cost threshold, for a stationary vehicle state ( The maximum acceptable cost under the condition of 0 is the benchmark parameter for threshold calculation, and its value needs to be calibrated through real vehicle scenario testing.

[0042] Speed-threshold coupling coefficient, used to adjust the influence of vehicle speed on the cost threshold. >0. The larger the value of , the more sensitive the cost threshold is to changes in vehicle speed.

[0043] The vehicle's current speed is obtained from the onboard CAN bus and needs to be processed by moving average filtering to eliminate signal jitter and ensure the stability of threshold calculation.

[0044] The activation speed limit for the low-speed search function is a fixed calibrated value (e.g., =10km / h), consistent with the function activation conditions.

[0045] By introducing a cost threshold that is dynamically adjusted in conjunction with vehicle speed, the system can adaptively relax or tighten the parking space selection criteria according to the real-time driving status of the vehicle. This makes the generation process of candidate parking spaces more environmentally adaptable, reduces the risk of rigid or unsuitable selection results caused by changes in vehicle speed, and improves the decision-making flexibility and reliability of the automatic parking system under different working conditions.

[0046] In one embodiment, the selection of the candidate parking space list based on the comparison result of the parking space selection cost and the cost threshold for each parking space includes: Based on the comparison between the parking space selection cost and the cost threshold for each parking space, parking spaces with a selection cost less than or equal to the cost threshold are selected. The selected parking spaces are sorted from lowest to highest cost, and one or more parking spaces with the lowest cost are selected to form the candidate parking space list.

[0047] For example, the parking space selection cost mentioned above is a quantitative indicator that comprehensively evaluates the ease of parking and the matching degree of user needs, while the cost threshold is a screening threshold dynamically calculated based on vehicle speed.

[0048] When generating the candidate parking space list, this application first compares the parking space selection cost of each parking space with the currently calculated cost threshold, filtering out all qualified parking spaces whose cost is less than or equal to the threshold. Then, these qualified parking spaces are sorted from smallest to largest according to their parking space selection cost, and one or more parking spaces with the lowest cost are selected to form the final candidate parking space list for subsequent route pre-planning. For example, when vehicles are traveling at low speeds, the system uses a lower cost threshold for strict screening, retaining only a few optimal parking spaces in the list; while when vehicle speeds are higher, the cost threshold is increased accordingly, allowing more parking spaces to pass the initial screening, and then the relatively better ones are selected through sorting.

[0049] By employing the aforementioned technical solution, through a progressive process of first performing threshold screening and then sorting by cost, the system can quickly focus on a batch of candidate targets from all available parking spaces that have both high parking suitability and meet the quantity management requirements under the current vehicle speed environment. This reduces the possibility of low decision-making efficiency or poor results caused by processing all parking spaces at once or rigid screening criteria, and lays a reasonable and efficient candidate foundation for subsequent fine-grained path planning.

[0050] S103. Based on the preset path planning process, perform path pre-planning for each parking space in the candidate parking space list to obtain the planning cost of each parking space in the candidate parking space list. The planning cost is used to evaluate the merits of the path from the current vehicle position to the corresponding parking space, and the parking space with the lowest planning cost is the target parking space.

[0051] For example, planning cost is a comprehensive indicator used to evaluate the quality of the planned path from the current vehicle position to the corresponding parking space. It reflects the execution efficiency and operational quality of the path by quantifying multiple characteristics of the path. The path pre-planning process refers to the process by which the system pre-simulates and generates a feasible parking path for each candidate parking space.

[0052] After obtaining the list of candidate parking spaces, this application will perform path pre-planning for each parking space in the list based on a preset path planning process. That is, it will simulate and generate a complete parking path from the vehicle's current position to the parking space and calculate the planning cost corresponding to the path. For example, the system will evaluate the geometric features and motion constraints of the path, generate a collision-free path that meets the vehicle's steering restrictions, and then comprehensively calculate the planning cost from multiple dimensions such as path length, shift frequency, trajectory smoothness, and the parking space's own adaptability. Finally, the parking space with the lowest planning cost will be determined as the target parking space, ensuring that the selected parking space not only has excellent static conditions, but also has the most efficient and reliable reachable path.

[0053] By employing the above technical solution, and by performing path pre-planning and calculating planning costs for each candidate parking space, the system can verify the feasibility and advantages of the path in advance during the decision-making stage. The dynamic execution cost of the path is incorporated into the parking space selection criteria, thereby reducing the risk of decision-making bias caused by only considering the static attributes of the parking space and ignoring the complexity of the actual parking path. This improves the scientific nature of the final target parking space selection and the overall parking efficiency of the automatic parking system.

[0054] In one embodiment, the preset path planning employs a hierarchical planning strategy, and the algorithm execution order of the hierarchical planning strategy is as follows: geometric splicing algorithm, hybrid A and RS curve algorithm, and hybrid A and RS segmentation optimization algorithm. The preset path planning process performs path pre-planning for each parking space in the candidate parking space list to obtain the planning cost for each parking space in the candidate parking space list, including: For each parking space in the candidate parking space list, the algorithm is called sequentially according to the algorithm execution order; If the algorithm planning for the current priority is successful, use its planning result and terminate the subsequent algorithm calls; If the algorithm planning of the current priority fails, the algorithm of the next priority is started to continue planning until the algorithm planning is successful; The planning cost of the currently calculated parking space is calculated based on the successful planning results of the algorithm.

[0055] For example, the geometric stitching algorithm mentioned above is an algorithm for fast path generation based on parking space geometric features and vehicle kinematic constraints; the hybrid A and RS curve algorithm is a fine path planning method that combines hybrid A-star global search and Reeds-Shepp curve local smoothing; the hybrid A and RS segmented optimization algorithm is a safety net planning strategy that allows for handling extremely complex scenarios through multiple gear shifts and segmented path stitching.

[0056] Regarding the first level, the geometric splicing algorithm: Upon entering the planning stage, the first-level planning algorithm is started by default. The geometric method has a very short computation time and can quickly calculate feasible paths. Algorithm principle: Based on the geometric features of the parking space (length, width, entrance angle) and the vehicle kinematic constraints (minimum turning radius), straight line segments and arc segments are directly spliced ​​to construct the initial path from the current position of the vehicle to the target pose of the parking space; then a closed-loop optimization strategy is introduced to fine-tune the key nodes of the initial path (such as turning points and parking space entrance points) to eliminate sudden changes in path curvature and ensure that the path meets the mechanical limit requirements for vehicle steering. Core advantages: The computational complexity of the geometric method is O(1), which has a very short computation time. It can achieve instantaneous response of "parking space recommendation - trajectory planning" in low-speed search scenarios and is suitable for more than 80% of simple parking scenarios (such as empty perpendicular parking spaces and parallel parking spaces without adjacent vehicles). Termination conditions: If the geometric stitching algorithm successfully generates a feasible path that satisfies the constraints, the trajectory is output directly, and the planning process of subsequent levels is terminated; if the path has a collision risk or violates kinematic constraints (such as insufficient turning radius), the second level of planning is triggered.

[0057] Regarding the second level, the hybrid A* and RS curve algorithm is a precise optimization strategy for complex scenarios. It is initiated after the first-level planning fails, aiming to generate a continuous and smooth trajectory that satisfies vehicle non-holonomic constraints. Algorithm principle: Hybrid A* global search: Using the rear axle center of the vehicle as the kinematic control point, a four-dimensional state space is constructed, including position, heading angle, and gear position. The search direction is guided by an Euclidean distance heuristic function, embedding vehicle kinematic constraints (minimum turning radius, gear shifting logic) during the search process to generate a discrete path node sequence from the starting point to the end point. RS curve local smoothing: Reeds-Shepp curves are interpolated and fitted between the discrete nodes generated by the hybrid A* algorithm. The RS curve, as the shortest path curve satisfying non-holonomic constraints, enables collision-free continuous transitions between nodes, ultimately outputting a smooth trajectory with continuous curvature that can be directly tracked. Time constraints and scenario adaptation: This algorithm has high computational complexity; to avoid impacting user experience, a strict upper limit on the planning time is required. ,in , For the number of candidate parking spaces, The number of parking directions can be selected for each parking space (parallel parking spaces only support rear-end parking, while perpendicular parking spaces support both front-end and rear-end parking). The exit direction corresponds one-to-one with the parking direction, including forward, left-front, and right-front exits (front-end parking) and forward, left-front, and right-front exits (rear-end parking). This algorithm is suitable for medium-complexity scenarios (such as perpendicular parking spaces with small distances between adjacent vehicles, and parallel parking spaces requiring precise obstacle avoidance). Termination condition: If in If a feasible smooth trajectory is generated internally, the trajectory is output; if all candidate parking space planning times out or the path is infeasible, the third-level planning is triggered.

[0058] Regarding the third level, the hybrid A* and RS segmented optimization algorithm: This layer serves as a safety net for extremely complex scenarios, activating after the second-level planning times out or fails. Its core objective is to prioritize the completion of the parking task, allowing for trajectory generation through "multiple gear shifts and segmented adjustments." Algorithm principle: Hybrid A* key node generation: Simplifies the state space dimension, searching only the key nodes of the trajectory (starting point, shift points, obstacle avoidance points, parking space target point), without pursuing path continuity between nodes, significantly reducing the computational load. RS curve segmented splicing: Calls a pre-calculated RS curve segment library (containing basic combination segments such as straight-arc and arc-arc), matches the optimal segment between adjacent key nodes, and completes the splicing; the algorithm explicitly allows multiple forward-reverse shifting operations, avoiding complex obstacles through segmented adjustments. Core Advantages and Time Constraints: This layer's algorithm sacrifices trajectory optimality for robust planning, making it suitable for extremely complex scenarios (such as narrow, angled parking spaces or irregularly shaped parking spaces surrounded by multiple obstacles). The generated trajectory is a segmented, continuous trajectory, which may involve multiple gear shifts. The path length and smoothness are slightly inferior to the previous two layers, but it ensures the vehicle successfully parks in the target space, achieving the "minimum guarantee of completion" functional goal. The upper limit for drawing time is relaxed to 2 seconds to ensure that a feasible trajectory can still be output in complex environments.

[0059] In the path pre-planning process, for each parking space in the candidate parking space list, the system sequentially calls the geometric stitching algorithm, the hybrid A and RS curve algorithm, and the hybrid A and RS segmented optimization algorithm for path planning. When an algorithm of a certain priority is called and successfully generates a feasible path, the system immediately uses the planning result of that algorithm as the output and terminates the calling of subsequent algorithms. If the current algorithm fails, the system automatically uses the algorithm of the next priority to continue trying to plan until an algorithm successfully generates a path. Finally, the planning cost of the parking space is calculated based on the successfully planned path result. For example, for a perpendicular parking space with open surroundings, the geometric stitching algorithm can usually quickly generate an effective path, and the system directly uses the result. However, for a narrow parking space with dense obstacles, geometric stitching may fail, and the system will use the hybrid A and RS curve algorithm for more refined search and smoothing. If it is still unsuccessful, the hybrid A and RS segmented optimization algorithm will be used to ensure the feasibility of the path.

[0060] The formula for calculating the planning cost is as follows:

[0061] : The planned path length. a>0, used to adjust the proportion of path length in the total cost. The larger the value of a, the more sensitive the algorithm is to path length, and the more it tends to generate short path trajectories. Physical meaning: Path length is directly related to the energy and time costs of parking. Shorter paths can effectively reduce user waiting time and meet customers' needs for efficient parking.

[0062] : Number of gear shifts (forward / reverse switching counts as 1). b>0, used to quantify the impact of gear shifting on the total cost. The larger the value of b, the more sensitive the algorithm is to frequent gear shifts, prioritizing trajectories with fewer or no gear shifts. Physical meaning: Frequent gear shifts reduce the smoothness of the parking process; reducing the number of gear shifts is the core optimization direction for improving parking comfort.

[0063] Path smoothness penalty, integral of the square of the trajectory curvature over the trajectory length domain, trajectory curvature Let be the instantaneous curvature along the trajectory length s, where ,( (where is the turning radius of the trajectory at that position). The total length of the trajectory; the integration interval is [0, ... This integral term covers the complete trajectory from the starting point to the ending point. `c > 0` adjusts the priority of trajectory smoothness in the total cost. A larger `c` value results in a stronger penalty for high-curvature segments, forcing a smooth change in trajectory curvature and avoiding sharp turns and abrupt curvature changes; a smaller `c` value tolerates local high-curvature segments to adapt to extreme scenarios such as narrow parking spaces. Physical meaning: This integral term is a classic quadratic form metric for evaluating trajectory geometric smoothness. A smaller integral value indicates a more uniform trajectory curvature distribution and better smoothness. A smooth trajectory reduces the control difficulty of the vehicle's steering motor, reduces tire-to-ground friction loss, and improves the accuracy and stability of trajectory tracking. Compared to the time-domain integral form, this length-domain integral does not rely on vehicle speed data, has higher computational robustness, and is consistent with the path length metric dimension, reducing the complexity of weight coefficient calibration.

[0064] The parking space selection cost is a comprehensive index calculated using the parking space selection cost formula (Equation 1), integrating multiple dimensions such as the relative distance between the parking space and the vehicle, heading angle deviation, parking space width adaptability, surrounding obstacle density, and user preferences. d>0 is used to balance the optimization priority between "parking space quality" and "trajectory characteristics." The larger the value of d, the more the algorithm tends to prioritize matching lower-quality parking spaces. Even if a parking space has a slightly higher trajectory length or number of gear shifts, a high-quality parking space can still be found; the smaller the value of d, the higher the priority of trajectory characteristics over parking space suitability. Physical meaning: Embedding the parking space selection cost into the trajectory planning cost function achieves joint optimization of "parking space selection - trajectory planning," avoiding the contradictory result of "optimal trajectory but poor parking space." Through coupling... This ensures that the final generated trajectory not only meets the vehicle's kinematic constraints, but also caters to the user's personalized parking needs (such as preference for charging station parking spaces, perpendicular parking spaces, etc.).

[0065] By employing the above technical solution and using a hierarchical planning strategy and calling algorithms in order of priority, the system can prioritize the use of simple algorithms with high computational efficiency while ensuring the success rate of planning, and only activate complex algorithms when necessary. This reduces the risk of planning delays or failures caused by the insufficient adaptability of a single algorithm, improves the overall efficiency and robustness of the path pre-planning process, and ensures the timeliness and reliability of candidate parking space evaluation.

[0066] Furthermore, in the parking stage, in addition to selecting the optimal parking space, this invention also designs a set of dynamic direction recommendation rules based on path planning complexity for parking space types that support multiple parking directions, in order to further improve parking efficiency. This rule is applied differently according to the different geometric characteristics of the parking space: for parallel parking scenarios, the system defaults to rear-end parking, as its path is relatively standard, so no direction recommendation is needed. For perpendicular parking and herringbone parking spaces, the system's default preferred direction is also rear-end parking; however, for further optimization, the system will calculate feasible paths for front-end parking in parallel and introduce a key judgment condition: if the number of segments of the planned front-end parking path plus a preset tolerance value (e.g., 2 segments) is still less than the number of segments of the rear-end parking path, then it is determined that under the current vehicle position and environment, the front-end parking path is simpler and more convenient to operate, and therefore, front-end parking is recommended as the better solution. Conversely, for reverse herringbone parking spaces, the system's default preferred direction is front-end parking. Its dynamic judgment logic is the opposite: if the number of segments in the planned rear-end parking path plus a preset tolerance value is less than the number of segments in the front-end parking path, the system will recommend rear-end parking. This rule, with "number of path segments" as the core comparison indicator, is designed to quantify the abstract "parking difficulty" into concrete path complexity. More path segments usually mean more steering and shifting operations are required. Therefore, the system intelligently and dynamically selects the parking direction that simplifies the parking operation, effectively avoiding path redundancy problems that may arise from a fixed default direction.

[0067] In one embodiment, the above method further includes: When the target vehicle automatically parks in the target parking space, the deviation between the actual driving trajectory of the target vehicle and the pre-planned reference trajectory is monitored based on the Hausdorf distance. If the Hausdorf distance exceeds a preset Hausdorf distance threshold, safety measures are triggered, including deceleration and replanning.

[0068] For example, the Hausdorf distance mentioned above is a metric used to quantify the overall shape difference between the actual driving trajectory of the target vehicle and the pre-planned reference trajectory. It can comprehensively reflect the global deviation of the trajectory. The preset Hausdorf distance threshold is a safety limit value pre-calibrated by the system to determine whether the trajectory deviation is within an acceptable range. Safety measures refer to the actions taken by the system when the deviation exceeds the limit, including controlling the vehicle to decelerate and initiating the trajectory replanning process.

[0069] During the automatic parking process of the target vehicle towards the target parking space, this application monitors the deviation between the actual driving trajectory and the reference trajectory in real time based on the Hausdorff distance, and compares the calculated Hausdorff distance with a preset Hausdorff distance threshold. If the Hausdorff distance is greater than the preset threshold, it indicates that the trajectory deviation has exceeded the safe range, and the system immediately triggers safety measures, such as issuing a deceleration command to reduce the vehicle speed and initiating a replanning process to regenerate a feasible parking trajectory based on the current vehicle position and environmental conditions. For example, when the vehicle's actual path gradually deviates from the reference trajectory due to changes in ground adhesion conditions or sensor noise, the system detects this accumulated deviation through the Hausdorff distance and intervenes in a timely manner to correct it, preventing the deviation from further expansion.

[0070] Specifically, during the automatic parking execution phase, the system first reuses the optimal trajectory generated in the pre-planning phase as the initial reference trajectory and sends it to the chassis control layer. This trajectory already satisfies the vehicle's kinematic constraints and has the optimal cost function. To address dynamic environmental disturbances and model mismatch during parking, the system introduces a real-time trajectory deviation monitoring and replanning triggering mechanism based on Hausdorff distance: using the rear axle center as the control point, the system collects the vehicle's actual pose sequence in real time and compares it with the corresponding pose sequence of the initial reference trajectory. The global morphological difference between the two is quantified by calculating the bidirectional Hausdorff distance. Compared to traditional single-point tracking errors (such as Euclidean distance), the core advantage of Hausdorff distance lies in its high sensitivity to overall trajectory offset. It can simultaneously and effectively identify local deviations caused by ground slippage and steering delay, as well as global deviations caused by dynamic obstacle intrusion and incorrect parking space boundary recognition, thereby providing a more comprehensive safety assessment. The system presets a dynamic Hausdorff distance threshold as a safety limit. This threshold is specifically calibrated based on the safety redundancy of the parking scenario: for open parking spaces with low obstacle density, the threshold can be appropriately relaxed to avoid minor deviations triggering unnecessary replanning; for narrow parking spaces with small distances between adjacent vehicles or complex environments with multiple dynamic obstacles, the threshold is strictly tightened to ensure parking safety. When the real-time calculated Hausdorff distance exceeds the preset threshold, the system immediately triggers multi-layered safety measures: First, it issues temporary control commands (such as deceleration or maintaining the current steering angle) to bring the vehicle into a relatively safe transient state; then, based on the vehicle's current actual position and environmental state, the system initiates a trajectory replanning process to quickly generate a new feasible trajectory; finally, the new trajectory smoothly replaces the original reference trajectory and is sent to the execution layer to continue completing the parking task. This closed-loop control mechanism of "tracking-monitoring-safe transient-replanning" effectively improves the system's response speed and handling capability to unexpected situations, ensuring the robustness and safety of the parking process.

[0071] Regarding the above plan 1. Initial Trajectory Reuse. The pre-planned trajectory is the optimal trajectory generated based on a three-layer programming algorithm. Its core characteristics are: satisfying vehicle kinematic constraints (minimum turning radius, non-integrity constraints); achieving the optimal cost function value, balancing path length, number of gear shifts, smoothness, and parking space adaptability; and strictly matching the parking direction selected by the user (e.g., parking with the front of the car in a perpendicular parking space, or parking with the rear of the car in a parallel parking space). When the system calls the initial reference trajectory, it simultaneously sends out the key parameters of the trajectory (curvature distribution, gear shift point position, target pose), providing a benchmark for steering and speed control in the chassis execution layer.

[0072] 2. Real-time path error calculation based on Hausdorff distance. Using the rear axle center of the vehicle as the control point for trajectory tracking, the actual pose sequence of the vehicle is acquired in real time. Simultaneously, extract the corresponding pose sequence of the initial reference trajectory. The core advantage of using bidirectional Hausdorff distance to calculate the deviation between two sets of pose sequences lies in its sensitivity to global trajectory deviation. Compared to traditional single-point tracking errors (such as the Euclidean distance between the current pose and the reference pose), it can simultaneously measure the overall morphological difference between the actual trajectory and the reference trajectory, effectively identifying two types of key deviations: local deviations: single-point pose shifts caused by vehicle slippage or steering delays; global deviations: overall trajectory shifts caused by dynamic obstacle intrusion or parking space boundary recognition errors. In parking scenarios, the value of the Hausdorff distance directly reflects the degree of deviation between the actual parking path and the planned path. The larger the value, the more significant the trajectory deviation, and the higher the risk of parking collisions.

[0073] 3. Replanning trigger conditions and execution strategy. The system pre-calibrates the Hausdorff distance threshold. Its value needs to be determined in conjunction with the safety redundancy of the parking scenario: for open parking spaces (low obstacle density), the threshold can be appropriately relaxed to avoid triggering unnecessary replanning due to minor deviations; for narrow parking spaces (small distance between adjacent vehicles) or complex environments (multiple dynamic obstacles), the threshold needs to be strictly tightened to ensure that the trajectory deviation is always within a safe range. During the parking process, the system calculates the Hausdorff distance at a fixed period (e.g., 100ms): if If the trajectory tracking is determined to be normal, continue executing the initial reference trajectory; if > The system immediately triggers the trajectory replanning process and simultaneously issues temporary control commands (such as deceleration or maintaining the current steering angle) to ensure the vehicle is in a safe state. After replanning is completed, the system smoothly replaces the original reference trajectory with the newly generated trajectory and sends it to the execution layer to continue the parking task, achieving closed-loop control of "tracking-monitoring-replanning".

[0074] This mechanism effectively addresses the environmental uncertainty during parking by monitoring global deviations in Hausdorff distance and replanning in real time. It avoids parking failures caused by minor deviations and corrects the trajectory in a timely manner in scenarios with large deviations, ultimately ensuring the robustness of the parking function and the driving safety of users.

[0075] By employing the above technical solutions, through real-time trajectory monitoring and deviation triggering mechanisms, the system can dynamically detect and respond to path tracking errors and environmental disturbances during the parking execution phase, reducing the possibility of parking interruptions or safety accidents caused by trajectory deviations or sudden obstacles, and improving the adaptability and reliability of the automatic parking process.

[0076] In one embodiment, the above method further includes: During the parking exit phase, the system recommends the optimal parking exit direction (left, center, right) to the user, improving parking efficiency and convenience. Specifically, for each available parking exit direction, the corresponding parking exit trajectory is pre-calculated. A cost is calculated for each pre-planned parking exit trajectory. The cost function can use the planning cost formula from the parking entry phase, or adjust the penalty parameters accordingly. The parking exit direction with the lowest calculated cost is recommended to the user, and the corresponding operation button is highlighted on the vehicle's infotainment interface.

[0077] This application extends the automatic parking function to the entire vehicle usage cycle, specifically targeting vehicle departure scenarios, and designs an independent intelligent recommendation scheme for parking exit directions. During the parking exit phase, the system provides users with optimal parking exit direction suggestions to improve departure efficiency and convenience. First, it identifies and predefines three basic parking exit directions: left, forward, and right. For each available parking exit direction, the system initiates a pre-planning process similar to the parking entry phase, simulating and calculating a complete and feasible parking exit trajectory from the current parking space to the departure target point for each direction. Subsequently, a comprehensive planning cost is calculated for each pre-planned parking exit trajectory to quantitatively evaluate the merits of that trajectory. This planning cost calculation can adopt the cost function from the parking entry phase, considering factors such as path length, number of gear shifts, and trajectory smoothness; simultaneously, to more accurately adapt to the characteristics of the parking exit scenario, the penalty parameter in this cost function can be appropriately adjusted. After completing the cost calculation for all directions, the system compares the results and selects the parking exit direction with the lowest planning cost, determining it as the optimal choice under the current circumstances. Finally, the optimal parking direction is recommended to the user through a human-machine interface (such as an in-vehicle display screen) using a highlighted button or other prominent method, guiding the user to trigger the process with a single click, thereby completing an efficient and smooth automatic parking operation. This solution effectively solves the problems of lack of pre-planning and blind decision-making in the parking phase of existing technologies, realizing full-process intelligentization of the parking function in both "entering" and "leaving" scenarios.

[0078] Furthermore, as a response to the above Figure 1 In addition to the implementation of the method shown, this embodiment of the invention also provides an automatic parking device for the above-mentioned... Figure 1 The method shown is implemented accordingly. This device embodiment corresponds to the foregoing method embodiment. For ease of reading, this device embodiment will not repeat the details of the foregoing method embodiment, but it should be clear that the device in this embodiment can implement all the contents of the foregoing method embodiment. Figure 2 As shown, the device includes: a calculation unit 21, a selection unit 22, and a planning unit 23, wherein... The calculation unit 21 is used to calculate the parking space selection cost of each available parking space for the target vehicle when a parking task is triggered, wherein the parking space selection cost is used to characterize the parking difficulty of the parking space relative to the target vehicle and the degree of matching with the user's needs. Selection unit 22 is used to select a candidate parking space list based on the comparison result of the parking space selection cost and the cost threshold for each parking space, wherein the cost threshold is determined based on the speed of the target vehicle; Planning unit 23 is used to perform path pre-planning for each parking space in the candidate parking space list based on a preset path planning process, so as to obtain the planning cost of each parking space in the candidate parking space list. The planning cost is used to evaluate the merits of the path from the current vehicle position to the corresponding parking space, and the parking space with the lowest planning cost is the target parking space.

[0079] The processor contains a kernel, which retrieves the corresponding program unit from memory. One or more kernels can be configured, and by adjusting kernel parameters, an automatic parking method can be implemented, addressing the problem of insufficient accuracy in parking space assessment and planning in existing automatic parking technologies.

[0080] This invention provides a computer-readable storage medium including a stored program that, when executed by a processor, implements the automatic parking method.

[0081] This invention provides a processor for running a program, wherein the program executes the automatic parking method during runtime.

[0082] This invention provides an electronic device, which includes at least one processor and at least one memory connected to the processor; wherein the processor is used to call program instructions in the memory to execute the automatic parking method described above. This invention provides an electronic device 30, such as... Figure 3 As shown, the electronic device includes at least one processor 301, and at least one memory 302 and bus 303 connected to the processor; wherein, the processor 301 and the memory 302 communicate with each other through the bus 303; the processor 301 is used to call program instructions in the memory to execute the above-mentioned automatic parking method.

[0083] The smart electronic devices mentioned in this article can be PCs, tablets, mobile phones, etc.

[0084] This application also provides a computer program product that, when executed on a process management electronic device, is suitable for executing a program that initializes the above-described automatic parking method steps.

[0085] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0086] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0087] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0088] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0089] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0090] This application also provides a computer program product, which includes computer software instructions that, when executed on a processing device, cause the processing device to perform actions such as... Figure 1The control flow of the memory in the corresponding embodiment.

[0091] A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0092] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0093] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0094] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0095] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0096] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0097] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An automatic parking method, characterized in that, include: When a parking task is triggered, the parking space selection cost for each available parking space for the target vehicle is calculated, wherein the parking space selection cost is used to characterize the ease or difficulty of parking the parking space relative to the target vehicle and the degree of matching with the user's needs. A list of candidate parking spaces is selected based on a comparison between the parking space selection cost and a cost threshold for each parking space, wherein the cost threshold is determined based on the speed of the target vehicle. Based on a preset path planning process, a path pre-planning is performed for each parking space in the candidate parking space list to obtain the planning cost of each parking space in the candidate parking space list. The planning cost is used to evaluate the merits of the path from the current vehicle position to the corresponding parking space, and the parking space with the lowest planning cost is the target parking space.

2. The method according to claim 1, characterized in that, The factors influencing the cost of parking space selection include: Distance assessment results, which are used to calculate the straight-line distance from the center of the rear axle of the vehicle to the center of the parking space entrance; Angle evaluation results, which are used to calculate the angle deviation between the vehicle's current driving direction and the parking space entrance direction; Width assessment results, which are used to assess parking spaces whose actual width is less than the standard width; Environmental assessment results, which are used to calculate the obstacle density around the parking space; Personalized assessment results are used to adjust the parking space selection cost based on vehicle status and user habits.

3. The method according to claim 2, characterized in that, When a parking task is triggered, the calculation of the parking space selection cost for each available parking space for the target vehicle includes: Obtain the distance assessment result, angle assessment result, width assessment result, environmental assessment result, and personalized assessment result for the current calculated parking space; Based on the distance assessment results, the angle assessment results, the width assessment results, the environmental assessment results, and the personalized assessment results, the parking space selection cost for the current calculated parking space is calculated using a predefined cost function.

4. The method according to claim 1, characterized in that, Also includes: The cost threshold is calculated based on the cost threshold, the current speed, the speed limit, and the speed coupling coefficient. The cost threshold represents the maximum acceptable cost when the vehicle is stationary. The speed limit is the speed limit that triggers the parking task. The speed coupling coefficient is used to adjust the influence of speed on the threshold. The cost threshold is directly proportional to the vehicle speed.

5. The method according to claim 1, characterized in that, The selection of the candidate parking space list based on the comparison result of the parking space selection cost and the cost threshold for each parking space includes: Based on the comparison between the parking space selection cost and the cost threshold for each parking space, parking spaces with a selection cost less than or equal to the cost threshold are selected. The selected parking spaces are sorted from lowest to highest cost, and one or more parking spaces with the lowest cost are selected to form the candidate parking space list.

6. The method according to claim 1, characterized in that, The preset path planning adopts a hierarchical planning strategy, and the algorithm execution order of the hierarchical planning strategy is as follows: geometric splicing algorithm, hybrid A and RS curve algorithm, and hybrid A and RS segmented optimization algorithm. The preset path planning process performs path pre-planning for each parking space in the candidate parking space list to obtain the planning cost for each parking space in the candidate parking space list, including: For each parking space in the candidate parking space list, the algorithm is called sequentially according to the algorithm execution order; If the algorithm planning for the current priority is successful, use its planning result and terminate the subsequent algorithm calls; If the algorithm planning of the current priority fails, the algorithm of the next priority is started to continue planning until the algorithm planning is successful; The planning cost of the currently calculated parking space is calculated based on the successful planning results of the algorithm.

7. The method according to claim 1, characterized in that, Also includes: When the target vehicle automatically parks in the target parking space, the deviation between the actual driving trajectory of the target vehicle and the pre-planned reference trajectory is monitored based on the Hausdorf distance. If the Hausdorf distance exceeds a preset Hausdorf distance threshold, safety measures are triggered, including deceleration and replanning.

8. An automatic parking device, characterized in that, Also includes: The calculation unit is used to calculate the parking space selection cost of each available parking space for the target vehicle when a parking task is triggered, wherein the parking space selection cost is used to characterize the parking difficulty of the parking space relative to the target vehicle and the degree of matching with the user's needs. A selection unit is used to select a list of candidate parking spaces based on a comparison between the parking space selection cost and a cost threshold for each parking space, wherein the cost threshold is determined based on the speed of the target vehicle. The planning unit is used to perform path pre-planning for each parking space in the candidate parking space list based on a preset path planning process, so as to obtain the planning cost of each parking space in the candidate parking space list. The planning cost is used to evaluate the merits of the path from the current vehicle position to the corresponding parking space, and the parking space with the lowest planning cost is the target parking space.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed by a processor, it implements the steps of the automatic parking method as described in any one of claims 1 to 7.

10. An electronic device, characterized in that, The electronic device includes at least one processor and at least one memory connected to the processor; wherein the processor is configured to invoke program instructions in the memory to execute the steps of the automatic parking method as described in any one of claims 1 to 7.

Images