Parking space management method, device and medium

Abstract

The application relates to the technical field of intelligent driving, in particular to a parking space management method, equipment and medium. When the method is used for automatic parking, parking space data around a target vehicle is acquired through environment perception information, and potential releasable parking spaces are quickly screened out based on preset geometric rules, while obviously unreleasable parking spaces are excluded. A verified parking space set after screening is taken as input, a path planning algorithm is called to verify the path feasibility of each parking space. The output result of the path planning algorithm is fed back to a user in real time through a vehicle-mounted display screen, so that the user can select a parking space. The scheme combines efficient screening of geometric rules and accurate verification of the path planning, and ensures the quick response and accuracy of the parking space release logic.

Description

Technical Field

[0001] This application relates to the field of intelligent driving technology, and in particular to a parking space management method, device and medium. Background Technology

[0002] With the rapid development of intelligent driving technology, automatic parking has become a standard feature in mid-to-high-end vehicles and is gradually penetrating into low-to-mid-end models. Parking space release is a crucial step in ensuring available parking spaces within an automatic parking system. Currently, there are three common methods for parking space release: geometric rule-based methods, which rely on preset geometric rules such as distance, angle, and size to filter parking spaces; geometric planning methods, which use fixed polylines and straight-line geometric path strategies to verify parking spaces; and search planning methods, which use a hybrid A* iterative search to generate paths and determine parking spaces.

[0003] However, geometric rule-based and geometric programming methods, due to their fixed rules or paths, cannot cover diverse scenarios and are prone to misjudgments. While search-based planning methods are highly adaptable, their computational efficiency is low, making it difficult to meet real-time requirements. Therefore, the above solutions struggle to achieve a balance between the accuracy and speed of parking space release. Summary of the Invention

[0004] This application provides a parking space management method, device, and medium to solve the aforementioned technical problems. This solution ensures that users can see available parking spaces in real time while reducing computation time through algorithm optimization, significantly improving the user experience and success rate of automatic parking functions.

[0005] In a first aspect, embodiments of this application provide a parking space management method, the method comprising:

[0006] Acquire environmental perception information of the target vehicle, wherein the environmental perception information includes at least one candidate parking space information;

[0007] The information of each candidate parking space is initially screened according to preset geometric rules to obtain a set of parking spaces to be verified.

[0008] A preset path planning algorithm is used to verify the feasibility of the path based on the set of parking spaces to be verified, so as to generate a releasable parking space identifier;

[0009] The releasable parking space identifier is sent to the vehicle display screen so that the user can select a parking space.

[0010] In one possible embodiment, the preset geometric rules include a first set of rules and a second set of rules, and the preliminary screening includes:

[0011] If all rules in the first rule set are satisfied, but no rule in the second rule set is satisfied, then the candidate parking space information is stored in the parking space set to be verified.

[0012] In one possible embodiment, preliminary screening includes:

[0013] If any rule in the first set of rules is not met, the candidate parking space information is stored in the non-release list.

[0014] If all rules in the first and second rule sets are satisfied, the candidate parking space information is stored in the release list.

[0015] In one possible embodiment, the set of parking spaces to be verified includes multiple parking space information entries. Route feasibility verification includes:

[0016] When any parking space information to be verified passes the path feasibility verification, a releaseable parking space identifier is generated based on that parking space information, and the verification of the remaining parking space information to be verified is stopped.

[0017] In one possible embodiment, it also includes:

[0018] The path planning task for multiple parking spaces to be verified is distributed to multiple computing units for parallel execution, and the results are synchronized through a shared memory mechanism.

[0019] In one possible embodiment, it also includes:

[0020] Based on obstacle data in environmental perception information, parking spaces containing preset obstacles are excluded from the parking space area.

[0021] Eliminate candidate parking spaces whose perceived confidence level is lower than a preset confidence threshold;

[0022] Remove parking spaces with incomplete information, such as those with missing corner points or abnormal parking space type identification.

[0023] In one possible embodiment, the preset geometric rules include a lateral distance threshold for the parking space entrance, a vehicle-parking space angle threshold, and further include:

[0024] The lateral distance threshold for parking space entrances is determined based on the obstacle density in the environmental perception information.

[0025] Determine the threshold for the angle between the parking space and the vehicle based on the type of parking space.

[0026] In one possible embodiment, it also includes:

[0027] The feasible paths that have passed the feasibility verification are subjected to dynamic verification and environmental real-time verification. The dynamic verification includes determining whether the path meets the minimum turning radius of the vehicle, and the environmental real-time verification includes determining whether there are any new obstacles on the path.

[0028] Secondly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0029] The memory stores the instructions that the computer executes;

[0030] The processor executes computer execution instructions stored in memory, causing the processor to perform the methods described above.

[0031] Thirdly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the methods provided above.

[0032] The parking space management method, equipment, and medium provided in this application uniformly receive the original perception data of parking spaces and obstacles output by surround view and radar fusion, providing a complete data source for subsequent parking space screening and path verification, realizing synchronous collection of multiple parking spaces, and supporting batch parallel processing of parking spaces. It quickly eliminates obviously unusable parking spaces by relying on low-computing-power geometric judgment, significantly reducing the number of parking spaces requiring subsequent path planning, reducing algorithm computation, and ensuring the basic response speed of the vehicle interface; only questionable parking spaces are retained for the next verification step, balancing screening efficiency. Non-standard and complex parking spaces that cannot be accurately determined by geometric rules are verified through path planning, compensating for the shortcomings of pure geometric rule scenario adaptation, ensuring that the final output parking spaces are all practically feasible for parking, and reducing the problem of erroneous parking space release. The verified valid parking spaces are visualized and displayed to the driver, clearly showing the available parking locations, completing the connection between algorithm decision-making and human-computer interaction, and enabling user-operable parking interaction. Attached Figure Description

[0033] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0034] Figure 1 An application scenario diagram of a parking space management method provided in this application;

[0035] Figure 2 A schematic flowchart illustrating a parking space management method provided in an embodiment of this application;

[0036] Figure 3 A schematic flowchart illustrating a parking space management method according to another embodiment of this application;

[0037] Figure 4 This is a schematic diagram of the structure of a parking space management device provided in an embodiment of this application;

[0038] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0039] The accompanying drawings have illustrated specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to specific embodiments. Detailed Implementation

[0040] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0041] It should be noted that in the embodiments of this application, certain software, components, models and other existing solutions in the industry may be mentioned. These should be regarded as exemplary and are only intended to illustrate the feasibility of implementing the technical solution of this application. However, they do not mean that the applicant has used or necessarily used the solution.

[0042] To clearly understand the technical solution of this application, the solutions of the prior art will be described in detail first.

[0043] With the continuous iteration of intelligent driving technology, automatic parking is gradually becoming more widespread, extending from high-end models to more affordable vehicles. Parking space availability, as a core decision-making step in automatic parking, directly determines the effectiveness of available parking spaces for users. Currently, the mainstream parking space availability methods fall into three categories: the first is the geometric rule method, which uses preset geometric thresholds such as distance, angle, and parking space size to initially screen parking spaces; the second is the geometric planning method, which uses standardized straight-line and polyline paths to verify parking feasibility; and the third is the search planning method, which uses a hybrid A* algorithm to iteratively search for a complete parking path to determine parking space availability. However, the first two methods rely on fixed judgment logic and path templates, which are prone to judgment errors in complex parking scenarios with various irregular shapes and obstacles. While the hybrid A* search planning method is adaptable to various complex scenarios, it suffers from high computational load and insufficient real-time performance. In summary, existing single solutions cannot simultaneously achieve both the accuracy of parking space availability judgment and the speed of interface response.

[0044] Therefore, when facing the technical problems in existing technologies, since traditional parking space release solutions only use geometric rules or search planning algorithms, they cannot simultaneously ensure real-time judgment and scene adaptation accuracy. Therefore, we first collect the vehicle's surrounding environment perception data output by vehicle surround view and radar fusion, batch-acquiring all candidate parking space information, fully covering all types of candidate parking spaces identified on the road surface, providing complete data support for batch unified verification. We then introduce preset geometric rules to conduct a lightweight preliminary screening of all parking spaces, quickly eliminating invalid parking spaces that clearly do not meet the basic parking conditions, leaving only parking spaces with questionable geometric conditions and difficult to directly determine, forming a set to be verified. This significantly reduces the computational volume of subsequent time-consuming path planning, ensuring the real-time response speed of parking space display on the vehicle interface. To solve the contradiction of pure geometric rules being prone to misjudgment in non-standard scenarios and the high latency of full computation of single search planning, we call path planning algorithms to perform parking feasibility verification on the set of parking spaces to be verified. Only parking spaces with effective collision-free paths are marked as releaseable, and finally pushed to the in-vehicle display screen for visualization, achieving intelligent parking space release management that balances rapid response and accurate reliability.

[0045] Figure 1 This is an application scenario diagram illustrating the parking space management method provided in this application, such as... Figure 1 As shown in the diagram, the parking space management method provided in this application corresponds to a scenario including: a first candidate parking space 101, a second candidate parking space 102, a third candidate parking space 103, and an autonomous vehicle 104. It is understood that the number of candidate parking spaces in the parking lot is unlimited; this embodiment only uses three parking spaces as an example for demonstration. The autonomous vehicle 104 integrates a parking space management device.

[0046] It should be noted that the autonomous vehicle 104 integrates devices for collecting environmental perception information, such as cameras.

[0047] Specifically, the parking space management device in the autonomous vehicle 104 acquires environmental perception information of the target vehicle, including information on the first candidate parking space 101, the second candidate parking space 102, and the third candidate parking space 103. Then, the parking space management device in the autonomous vehicle 104 performs preliminary screening of the candidate parking space information according to preset geometric rules to obtain a set of parking spaces to be verified. Next, a preset path planning algorithm is used to verify the feasibility of the path based on the set of parking spaces to be verified, generating a releasable parking space identifier. Finally, the parking space management device sends the releasable parking space identifier to the onboard display screen of the autonomous vehicle 104 for the user to select a parking space.

[0048] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0049] Figure 2 This is a flowchart illustrating a parking space management method provided in an embodiment of this application, as shown below. Figure 2 As shown, the executing entity in this embodiment is a parking space management device. This device can be implemented through a computer program, or through a medium storing the relevant computer program, such as a USB flash drive and / or optical disc, or through an electronic device integrating or installing the relevant computer program, such as a chip, server, or server cluster. The parking space management method provided in this embodiment includes the following steps:

[0050] S201. Obtain environmental perception information of the target vehicle, including at least one candidate parking space.

[0051] The target vehicle refers to the vehicle itself, which is equipped with an automatic parking system and is currently performing a parking space search process.

[0052] Among them, environmental perception information refers to the comprehensive data of the vehicle's surrounding environment output after being fused and processed by the vehicle's surround-view cameras, ultrasonic radar, millimeter-wave radar, etc. The information may include road edge lines, parking space markings, obstacle outline coordinates, ground locks, wheel chocks, vehicle positioning, vehicle heading angle, etc.

[0053] Among them, the information of the parking space to be selected refers to the information of a single parking space, such as the global coordinates of the four corners of the parking space, the length and width of the parking space, the orientation of the parking space entrance, the location of the wheel chocks, the markings of obstacles inside the parking space, the perception confidence level, the parking space type and other geometric and attribute parameters.

[0054] It is understandable that the number of parking spaces identified within the vehicle's field of vision corresponds to the number of independent sets of candidate parking space information generated.

[0055] Specifically, once the vehicle activates its automatic parking search function and its speed is below a certain threshold, all onboard sensing hardware continuously and synchronously collects images and distance measurement data from the vehicle's surroundings. In real time, the raw data from multiple sensors undergoes spatiotemporal alignment, coordinate unification, and noise filtering. Complete parking space markings are identified on the road surface, and the rectangular outline of the parking space is fitted based on the intersection points of the markings. The spatial features and internal obstacle status of each parking space are extracted one by one to generate independent candidate parking space information.

[0056] Furthermore, all identified candidate parking space information is packaged together and integrated with other environmental data such as roads and obstacles into a complete environmental perception information data package, which is continuously transmitted to the parking space management device in real time.

[0057] Optionally, this step is dynamically updated throughout the vehicle's movement. As the vehicle moves forward, the camera and radar continuously refresh the field of view, and newly identified parking spaces are added to the environmental perception information in real time, while parking spaces that have moved out of the field of view are automatically removed.

[0058] S202. Perform preliminary screening of each candidate parking space information according to preset geometric rules to obtain a set of parking spaces to be verified.

[0059] Among them, the preset geometric rules refer to a set of spatial judgment constraint standards that have been pre-calibrated and solidified in the parking space management module, and are constructed based on planar geometric relationships. This may include, for example, a first layer of hard basic geometric rules used to verify basic parking conditions such as vehicle observation conditions, parking space physical dimensions, relative positions of vehicles and parking spaces, and parking space orientation matching. A second layer of traffic diversion verification rules includes anti-shake time intervals, vehicle displacement thresholds, and parking space entrance passage width constraints.

[0060] Among them, the parking space set to be verified refers to the parking space data group formed separately after batch screening by geometric rules. It cannot be directly determined as a releaseable parking space and needs to be handed over to the path planning algorithm for further verification of parking feasibility.

[0061] Specifically, after receiving all the parking space information, each parking space data is individually evaluated using a set of preset geometric rules. First, the first layer of hard, fundamental geometric rules is traversed, sequentially checking whether the vehicle has passed the parking space's reference corner point, the lateral and longitudinal distances between the vehicle and the parking space entrance, the angle between the vehicle's heading and the parking space entrance, the orientation of the parking space's wheel chocks, and whether the parking space's length and width dimensions meet the vehicle's passage allowance. If any of these fundamental geometric rules fails to meet the criteria, the parking space is directly classified into the non-releaseable group and will not participate in subsequent evaluations.

[0062] Optionally, if the parking space passes all the first-level hard rules completely, the second-level diversion verification rules are then executed, and the anti-shake time interval of the parking space release, the displacement difference between the two recognitions of the parking space by the vehicle, and the passage width of the parking space entrance passage are checked in turn.

[0063] Furthermore, parking spaces that satisfy all second-level rules are directly identified as available parking spaces and stored in the releaseable group. Parking spaces that do not satisfy any second-level rule are uniformly collected and organized to form a structured set of parking spaces to be verified, which is then fully output to the downstream path planning module for subsequent parking path feasibility verification.

[0064] Optionally, the preset geometric rules may include: a) determining whether the vehicle has passed a specified corner point of the parking space. b) determining whether the lateral distance between the vehicle and the parking space entrance exceeds a specified threshold, such as within [0.3m, 4.5m]. c) determining whether the longitudinal distance between the vehicle and the parking space exceeds a specified threshold, such as 15m. d) determining whether the wheel chock position and parking direction meet the requirements. e) determining whether the angle between the vehicle's angle and the parking space's orientation meets the requirements, such as requiring the angle between the vehicle's orientation and the parking space entrance's orientation to be below 60 degrees. f) determining whether the length of the target parking space meets the requirements, such as for perpendicular parking spaces, parking space length - threshold 1, for horizontal parking spaces, parking space length + threshold 2. g) determining whether the width of the target parking space meets the requirements. h) determining whether the time interval between the current time and the previous frame's parking space release time meets the requirements. i) determining whether the current position and the position when the parking space was released in the previous frame meet the requirements. j. Determine if the width of the parking space entrance channel meets the requirements. For example, determine if there are obstacles in a certain area of ​​the parking space entrance based on the parking space type. If there are no obstacles, it is considered that the width of the vehicle parking channel is met and can be released for users to choose. If there are obstacles, a path planning will be attempted to determine whether it can be released.

[0065] S203. A preset path planning algorithm is used to verify the feasibility of the path based on the set of parking spaces to be verified, so as to generate a releasable parking space identifier.

[0066] Among them, the preset path planning algorithm refers to a hybrid A-star search planning algorithm that is pre-integrated and adapted to vehicle kinematic constraints, and then subjected to lightweight improvements.

[0067] Among them, the digital path feasibility verification process simulates the complete parking action of a vehicle within the set of parking spaces to be verified, deduces the entire driving trajectory, and verifies whether the trajectory simultaneously meets all parking constraints such as vehicle movement restrictions, no obstacles in the surrounding area, and sufficient passage space.

[0068] Among them, the releasable parking space identifier refers to the exclusive marking information assigned to a parking space that has been fully verified and confirmed to be able to be parked normally. It includes the unique ID of the parking space, the space parameters, the recommended starting and ending coordinates of parking, and other supporting information.

[0069] Specifically, the set of parking spaces to be verified generated in the previous step is read, and each candidate parking space is retrieved from the set for individual feasibility verification. First, based on the current vehicle location, vehicle size, and surrounding obstacle perception data, a local drivable grid area is constructed with the target parking space as the endpoint, defining the permitted passage range for the vehicle, and marking all obstacles inside the parking space and in the passageway as impassable areas. Then, a lightweight hybrid A* algorithm is called to perform path search, strictly following the kinematic constraints such as the vehicle's front wheel steering limits and minimum turning radius to expand the driving nodes outward, continuously matching a continuous driving trajectory from the vehicle's current position into the parking space.

[0070] Optionally, during the search process, the safe distance between trajectory nodes and obstacles and road boundaries is detected in real time. Once a complete parking trajectory that is collision-free and meets vehicle control restrictions is generated, the algorithm iteration is immediately stopped and the search for a better path is no longer continued.

[0071] Furthermore, if a valid parking trajectory is successfully found, a unique releaseable parking space identifier is generated for that parking space. If a compliant driving path cannot be generated after a thorough search, the parking space is classified into a non-releaseable group, and no corresponding identifier is generated.

[0072] S204. Send the releasable parking space identifier to the vehicle display screen so that the user can select a parking space.

[0073] Specifically, the system collects all parking space data with releasable parking space labels, performs data format adaptation and conversion, transforms the parking space coordinates, dimensions, and location information within the labels into visual rendering data recognizable by the display screen, filters out invalid data corresponding to non-releasable parking spaces, and retains only the label information of compliant parking spaces. After completing the format conversion, the system batch pushes out the visual data accompanying all releasable parking space labels.

[0074] Understandably, after receiving data, the in-vehicle display combines it with real-time vehicle location and road marking rendering to highlight and display parking spaces with releasable parking space icons in a selectable style on the top-down view interface, distinguishing between unselectable grayed-out invalid parking spaces. The driver can directly click on any highlighted parking space on the screen to confirm the parking target, and the display simultaneously sends the selected vehicle ID back to the planning and control module to initiate the automatic parking path planning and vehicle control process.

[0075] Optionally, if the set of parking spaces to be verified has no valid parking spaces after path verification, the display screen will only show the road environment and will not render any available parking space graphics.

[0076] The parking space management method provided in this application uniformly receives the original perception data of parking spaces and obstacles output by surround view and radar fusion, providing a complete data source for subsequent parking space screening and path verification, realizing synchronous collection of multiple parking spaces and supporting batch parallel processing of parking spaces. It quickly eliminates obviously unusable parking spaces using low-computing-power geometric judgment, significantly reducing the number of parking spaces requiring subsequent path planning, reducing algorithm computation, and ensuring the basic response speed of the vehicle interface; only questionable parking spaces are retained for the next verification step, balancing screening efficiency. Non-standard and complex parking spaces that cannot be accurately determined by geometric rules are verified through path planning, compensating for the shortcomings of pure geometric rule-based scenario adaptation, ensuring that the final output parking spaces are all practically feasible for parking, and reducing the problem of erroneous parking space release. The verified valid parking spaces are visualized and displayed to the driver, clearly showing the available parking locations, completing the connection between algorithm decision-making and human-computer interaction, and enabling user-operable parking interaction.

[0077] As an optional implementation, based on the above embodiments, the preset geometric rules include a first set of rules and a second set of rules, and the preliminary screening includes:

[0078] If all rules in the first rule set are satisfied, but no rule in the second rule set is satisfied, then the candidate parking space information is stored in the parking space set to be verified.

[0079] The first set of rules refers to the underlying hard geometric judgment criteria, which includes basic geometric constraints such as vehicle observation conditions, relative space between the vehicle and the parking space, the size of the parking space itself, and the orientation matching of the parking space. It is the minimum physical threshold for a parking space to complete parking.

[0080] The second set of rules refers to the anti-shake and traffic channel auxiliary constraint standards, which include verification items such as parking space release time interval, vehicle displacement distance, and parking space entrance channel width.

[0081] Specifically, each candidate parking space is read sequentially. First, all judgment conditions within the first rule set are fully traversed, and each condition is verified against hard geometric requirements such as the vehicle's field of vision, the lateral and longitudinal distance between the vehicle and the parking space, the angle between the vehicle and the parking space, the length and width of the parking space, and wheel well matching. If any of the first rule conditions for a parking space is not met, it is directly determined as a completely invalid parking space, diverted to the non-releaseable group, and no further checks are performed. If a parking space passes all the judgment conditions in the first rule set, it proceeds to the second rule set verification process.

[0082] Furthermore, the three constraints within the second set of rules are checked sequentially: the time interval between two parking space releases, the difference in vehicle displacement, and the width of the parking space entrance passage. If any condition in the second set is not met, it indicates that the parking space is subject to sensory jitter interference or the passage space is questionable, and it cannot be directly determined as a releaseable parking space. This parking space information is then collected and stored in the set of parking spaces to be verified, awaiting subsequent route planning verification. If all the second rules are met, the space is directly classified into the releaseable group without needing to proceed to the route verification stage.

[0083] The parking space management method provided in this application separates the minimum physical threshold for parking from anti-shake and channel auxiliary constraints, distinguishing between two types of parking space problems: absolutely impossible parking and questionable parking spaces awaiting verification. This avoids a uniform standard for one-size-fits-all screening, compensates for the shortcomings of pure geometric rules in handling non-standard scenarios, and improves the accuracy of parking space release determination. The method only verifies the second rule after the first rule is applied, reducing meaningless auxiliary rule judgments. For parking spaces where the basic conditions are completely invalid, the second rule verification is skipped, further reducing the overall vehicle computing power consumption.

[0084] As an optional implementation method, based on the above embodiments, the preliminary screening includes:

[0085] If any rule in the first set of rules is not met, the candidate parking space information is stored in the non-release list.

[0086] If all rules in the first and second rule sets are satisfied, the candidate parking space information is stored in the release list.

[0087] The "non-release list" refers to a group of data that includes parking spaces whose basic geometric conditions are completely invalid. Parking spaces in the list do not need to be included in subsequent path planning and verification; they are directly determined not to be displayed to the user.

[0088] The releaseable list refers to a group of parking spaces that have passed all double-layer rule verifications. Parking spaces in the list do not require route planning verification and can be directly pushed to the vehicle display screen as available parking spaces.

[0089] Specifically, each candidate parking space is read one by one, and all judgment conditions in the first rule set are verified item by item first. During the traversal of the first rule, if any hard geometric condition of the parking space fails to meet the standard, the parking space information is immediately classified and stored in the non-release list, all remaining verification steps for the parking space are terminated, and the second rule judgment is no longer executed.

[0090] Optionally, if the parking space passes all the judgment items in the first rule set, the verification of all constraints in the second rule set continues in sequence. If every constraint in the second rule is satisfied, it means that the parking space has no perceptible vibration and the passage space is sufficient, and the parking space information is directly stored in the release list.

[0091] The parking space management method provided in this application relies on lightweight geometric calculations to quickly eliminate parking spaces that do not meet the physical conditions for parking, directly preventing invalid parking spaces from entering subsequent complex calculation processes. This significantly reduces the number of parking spaces that need to be processed in path planning, lowers onboard computing power consumption, and ensures the real-time display of parking spaces on the vehicle interface. If both layers of rules are satisfied, the spaces are stored in a release list, eliminating the need to call time-consuming path planning algorithms and allowing for direct and rapid display on the screen, while also considering interface responsiveness. Simultaneously, the dual rule verification filters out false pop-up issues caused by perceived jitter and narrow passageways, ensuring the stability and reliability of directly allowed parking spaces and reducing the unpleasant experience of user interface flickering and invalid parking space pop-ups.

[0092] As an optional implementation, based on the above embodiments, the set of parking spaces to be verified includes multiple parking space information to be verified, and the route feasibility verification includes:

[0093] When any parking space information to be verified passes the path feasibility verification, a releaseable parking space identifier is generated based on that parking space information, and the verification of the remaining parking space information to be verified is stopped.

[0094] Among them, the parking space information to be verified refers to the complete structured data of a single parking space in the set, including all the basic parameters used for path planning such as parking space coordinates, size, entrance orientation, channel parameters, current vehicle positioning and attitude, etc. Each piece of information corresponds to an independent candidate parking space.

[0095] Specifically, the complete set of parking spaces to be verified is read, and individual parking space information within the set is retrieved sequentially in a preset order to verify the feasibility of the path. For the currently retrieved parking space, a locally passable area is constructed by combining real-time vehicle posture and surrounding obstacle perception data. A lightweight search planning algorithm is then called to deduce the complete parking trajectory, while simultaneously verifying whether the trajectory avoids obstacles and meets motion constraints such as the vehicle's minimum turning radius.

[0096] Furthermore, if a collision-free and compliant complete parking path is successfully derived from the simulation for this parking space, a corresponding releaseable parking space identifier is immediately generated based on all parameters of that parking space. After generating the identifier, the planning and verification process for all remaining parking spaces in this set is terminated directly, and the remaining unverified parking spaces are no longer traversed or calculated.

[0097] Optionally, if the current parking space route calculation fails, no corresponding identifier is generated, and the next parking space information to be verified in the set is retrieved and the route verification is repeated until the first successfully planned parking space is found, or all parking spaces in the set have been verified.

[0098] The parking space management method provided in this application generates a releaseable parking space identifier as soon as the first parking space is verified, without waiting for all questionable parking spaces to complete the calculation. This significantly shortens the overall time from parking space identification to interface display, greatly improves the real-time performance of parking space display, and enhances the user's waiting experience when searching for parking spaces. It proactively discards subsequent invalid planning calculations, reducing the number of algorithm iterations and lowering computational resource consumption from the source. While ensuring the provision of valid parking spaces, it maximizes the compression of path planning computational overhead, making it suitable for vehicle hardware platforms with limited computing power.

[0099] As an optional implementation, based on the above embodiments, it further includes:

[0100] The path planning task for multiple parking spaces to be verified is distributed to multiple computing units for parallel execution, and the results are synchronized through a shared memory mechanism.

[0101] Among them, the parking spaces to be verified refer to candidate parking spaces that meet all the hard basic geometric rules after the initial screening by the hierarchical geometric rules, but have doubts in the second rule verification and cannot be directly released. The parking feasibility needs to be verified through path planning.

[0102] The path planning task refers to the entire set of calculations for a single parking space to be verified, including complete parking trajectory simulation, collision detection, and kinematic constraint verification. Each task corresponds to the feasibility judgment of only one parking space.

[0103] The computing unit refers to the independent hardware processing cores within the vehicle processor that can perform operations synchronously. Each unit has independent computing resources and can process computing tasks in parallel without interfering with each other.

[0104] The shared memory mechanism refers to the public data storage area reserved by the vehicle system. All computing units can read and write data in this area, enabling real-time communication of calculation results and basic parking space information between different processing kernels without the need for additional data relay transmission.

[0105] Specifically, after completing the initial screening of geometric rules and generating a set of parking spaces to be verified containing multiple parking space data, the path planning work corresponding to each parking space in the set is split into independent single-parking-space path planning tasks. Based on the number of idle computing units in the current onboard processor, the split planning tasks are assigned to different idle computing units one by one, and multiple computing units simultaneously start the path deduction work of their respective assigned parking spaces.

[0106] Understandably, during the computation process of all computing units, common basic data is uniformly read from shared memory, including real-time vehicle positioning, vehicle motion parameters, a global obstacle grid map, and the geometric features of each parking space. After each computing unit completes the path feasibility verification for its own parking space, it immediately writes the planning result for that parking space into the corresponding storage partition of shared memory.

[0107] Furthermore, the system continuously polls the shared memory for updated parking space verification results, collects the judgment information output by all computing units, distinguishes between successfully planned and unplanned parking spaces, and generates batch releaseable parking space identifiers for all verified parking spaces.

[0108] The parking space management method provided in this application splits and distributes multi-parking space planning tasks to multiple computing units for parallel execution, breaking the computational limitation of single-core serial verification. Multiple parking spaces simultaneously perform path searches, significantly reducing the total verification time of the entire set of parking spaces to be verified and improving the overall response speed of parking space release. Shared memory provides lower data read / write latency, and each computing unit can read unified vehicle and environmental baseline data in real time, ensuring that different units have completely consistent judgment standards for the same road and vehicle parameters, avoiding judgment bias caused by data asynchrony between multiple units. The parking space planning results produced by each unit can be instantly aggregated, reducing the additional computing power overhead caused by data interaction.

[0109] As an optional implementation, based on the above embodiments, it further includes:

[0110] Based on obstacle data in environmental perception information, parking spaces containing preset obstacles are excluded from the parking space area.

[0111] Eliminate candidate parking spaces whose perceived confidence level is lower than a preset confidence threshold;

[0112] Remove parking spaces with incomplete information, such as those with missing corner points or abnormal parking space type identification.

[0113] Among them, obstacle data refers to the structured information of road obstructions, which includes the spatial coordinates and occupied area of ​​objects such as ground locks, cones, other vehicles, curbs, and guardrails.

[0114] Pre-defined obstacles refer to objects that are marked and determined in advance to obstruct vehicles from parking, including vehicles parked in parking spaces, parking space locks, large cones, fixed barriers, etc.

[0115] Among them, the perception confidence score refers to the confidence score output by the perception module, which represents the accuracy of recognizing the outline and markings of the current parking space. The higher the value, the more reliable the parking space recognition result.

[0116] Among them, the pre-set confidence threshold refers to the pre-calibrated fixed confidence threshold value, which serves as the criterion for judging whether the parking space recognition result is valid.

[0117] Among them, the corner point of a parking space refers to the intersection of the four marking lines that form the rectangular outline of the parking space.

[0118] Among them, parking space type recognition anomaly refers to the mismatch between the identified parking space category and geometric features. For example, the size meets the requirements for a vertical parking space but it is identified as a side parking space, or the outline cannot be classified as a standard side, vertical, or oblique parking space.

[0119] Specifically, before the initial screening using geometric rules, complete environmental perception information is read uniformly, and obstacle datasets and all candidate parking space information are extracted. A three-layer elimination process is then performed sequentially. The first layer iterates through each candidate parking space, extracting the coordinates of the complete occupied area of ​​the parking space, and comparing this coordinates with the spatial extent of all obstacles. If the obstacle area falls within the parking space and belongs to the preset obstacle category, the parking space is directly eliminated and will not participate in any subsequent processing.

[0120] Furthermore, in the second layer, the perception confidence score corresponding to each of the remaining parking spaces is read and compared with a preset confidence threshold. If the parking space confidence score is lower than the critical threshold, it means that the parking space markings are blurry and the outline is at risk of distortion, and the data for that parking space is directly removed.

[0121] Furthermore, the third layer verifies the completeness of the remaining parking spaces' features. Firstly, it checks whether the coordinates of all four corner points of the parking space have been identified and collected; if any corner point is missing, it is considered incomplete data. Secondly, it matches the parking space's geometric dimensions with the identified parking space type. If the size features and label type do not match, and the space cannot be classified as a standard parking space, it is considered an abnormal parking space type identification. Parking spaces with missing corner points or abnormal types are uniformly removed.

[0122] Understandably, after all three layers of elimination are completed, the remaining compliant parking spaces are collectively collected and used as input data for the subsequent preliminary screening using geometric rules. The entire pre-elimination process is repeated after each frame of perception data is updated.

[0123] The parking space management method provided in this application eliminates parking spaces occupied by vehicles, parking locks, or other obstacles, reducing redundant calculations and preventing accidental release of parking spaces. It removes parking spaces with low confidence or distorted recognition, mitigating errors caused by recognition mistakes. It also removes incomplete parking spaces with missing corners or incorrect type matching, ensuring the normal operation of subsequent geometric calculations and planning processes. This pre-filtering mechanism significantly reduces the number of parking spaces to be processed, lowers onboard computing power consumption, improves the reliability of parking space recognition, and maintains real-time response speed for parking space searching.

[0124] As an optional implementation, based on the above embodiments, the preset geometric rules include a lateral distance threshold for the parking space entrance, a threshold for the angle between the vehicle and the parking space, and also include:

[0125] The lateral distance threshold for parking space entrances is determined based on the obstacle density in the environmental perception information.

[0126] Determine the threshold for the angle between the parking space and the vehicle based on the type of parking space.

[0127] Among them, the lateral distance threshold at the parking space entrance refers to the critical value of the effective distance between the vehicle and the parking space in the direction perpendicular to the edge of the parking space entrance, which is used to determine whether the vehicle is in a reasonable lateral range that is identifiable and can be parked.

[0128] Among them, the vehicle-parking space angle threshold refers to the maximum allowable angle between the vehicle's own driving direction and the parking space entrance orientation, which is used to determine whether the vehicle's posture is suitable for parking operations.

[0129] Among them, obstacle density refers to the density of obstacles distributed in the passage area in front of the parking space entrance.

[0130] Among them, parking space type refers to the category of parking space that is identified and classified, including three categories: parallel parking space, perpendicular parking space, and angled parking space.

[0131] Specifically, real-time environmental perception information is read, all obstacle data within the parking space entrance passage area is extracted, the coverage and quantity of obstacles within the area are statistically analyzed, and the current obstacle density value of the passage is calculated. Based on a preset mapping relationship, a lateral distance threshold for the parking space entrance adapted to the current road conditions is obtained. The denser the obstacles in the passage, the smaller the effective lateral distance range will be, avoiding misjudgments due to obstacle obstruction when the vehicle is too far from the parking space.

[0132] Furthermore, the identification type of the parking space to be verified is read, and according to the parking steering characteristics of lateral, perpendicular, and angled parking spaces, a corresponding exclusive vehicle-space angle threshold is matched; lateral parking spaces allow for a larger vehicle body offset angle, while perpendicular parking spaces have stricter requirements for vehicle body alignment angle.

[0133] Understandably, after completing the dynamic assignment of the two types of thresholds, the real-time adapted thresholds are used to verify the lateral distance between the vehicle and the parking space and the angle between the vehicle and the parking space, in order to determine whether the parking space passes the first set of rules. The entire dynamic threshold update process is executed synchronously with each frame of perceived data.

[0134] The parking space management method provided in this application breaks away from the limitation of the traditional geometric rule threshold being unchanging. It can adapt to the current road and parking space scenario, greatly reduce the probability of false filtering of geometric rules in complex scenarios, and improve the accuracy of parking space release determination.

[0135] As an optional implementation, based on the above embodiments, it further includes:

[0136] The feasible paths that have passed the feasibility verification are subjected to dynamic verification and environmental real-time verification. The dynamic verification includes determining whether the path meets the minimum turning radius of the vehicle, and the environmental real-time verification includes determining whether there are any new obstacles on the path.

[0137] Among them, a feasible path refers to a complete driving trajectory that, through path planning and deduction, meets the basic space collision-free constraints and is preliminarily determined to be able to enter the parking space.

[0138] The minimum turning radius refers to the minimum turning radius that the vehicle body can complete when the front wheels are fully turned to their steering limit. It is an inherent motion parameter of the vehicle.

[0139] Specifically, after a preliminary feasible path is generated through path feasibility verification for the parking space, a two-layer verification process is initiated simultaneously. First, a dynamics verification is performed, extracting the trajectory turning curvature segment by segment along the entire feasible path and comparing the turning radius corresponding to each segment with the vehicle's inherent minimum turning radius. If the required turning radius for any segment is less than the vehicle's limit turning radius, it means the vehicle cannot complete the turning maneuver, and the path is directly deemed invalid.

[0140] Furthermore, real-time environmental verification is performed by retrieving the latest real-time environmental perception information, extracting the obstacle distribution across the entire path coverage area, comparing it with the obstacle data from the previous planning round, and checking for any new obstacles in the path lanes, parking space entrances, and inside the parking spaces. If any new obstacle appears in any area of ​​the path, the path is deemed to have a collision risk and is deemed invalid.

[0141] Understandably, only when both dynamics verification and environmental real-time verification pass can the parking space be confirmed as fully compliant, and a corresponding releaseable parking space identifier be generated. If either verification fails, the feasible path is discarded, and the parking space is determined to be unreleaseable.

[0142] The parking space management method provided in this application improves the accuracy of parking space release judgment by matching vehicle steering limits to filter trajectories that are theoretically feasible but practically impossible to achieve. This avoids releasing seemingly usable parking spaces to users that are actually impossible to navigate into, thus improving the reliability of parking space release decisions. While path planning is based solely on historical perception data, roads are subject to dynamic changes such as pedestrians and temporary obstacles. Real-time verification can capture obstacles that suddenly appear after planning is completed, eliminating the lag caused by static planning, avoiding parking collision risks, and improving driving safety.

[0143] Figure 3 A schematic flowchart of a parking space management method provided in another embodiment of this application is shown below. Figure 3 As shown, the parking space management method provided in this embodiment includes the following steps:

[0144] S301. Obtain environmental perception information of the target vehicle, including at least one candidate parking space.

[0145] S302. Based on obstacle data in the environmental perception information, remove parking spaces in the parking area that include preset obstacles.

[0146] S303. Eliminate candidate parking spaces whose perceived confidence level is lower than the preset confidence threshold.

[0147] S304. Remove parking spaces with incomplete information, such as those with missing corner points or abnormal parking space type identification.

[0148] S305. Determine the lateral distance threshold of the parking space entrance based on the obstacle density in the environmental perception information.

[0149] S306. Determine the threshold for the angle between the parking space and the vehicle based on the parking space type.

[0150] S307. Perform preliminary screening of each candidate parking space information according to preset geometric rules. The preset geometric rules include a first set of rules and a second set of rules.

[0151] S308. If all rules in the first rule set are satisfied, but no rule in the second rule set is satisfied, then the candidate parking space information is stored in the parking space set to be verified.

[0152] S309. If any rule in the first rule set is not satisfied, the candidate parking space information shall be stored in the non-release list.

[0153] S310. If all rules in the first rule set and the second rule set are satisfied, then the candidate parking space information is stored in the release list.

[0154] S311. Distribute the path planning tasks for multiple parking spaces to be verified to multiple computing units for parallel execution, and synchronize the results through a shared memory mechanism.

[0155] S312. When any parking space information to be verified passes the path feasibility verification, a releaseable parking space identifier is generated based on the parking space information to be verified, and the verification of the remaining parking space information to be verified is stopped.

[0156] S313. Perform dynamic verification and environmental real-time verification on feasible paths that have passed feasibility verification. Dynamic verification includes determining whether the path meets the minimum turning radius of the vehicle, and environmental real-time verification includes determining whether there are any newly appearing obstacles on the path.

[0157] S314. Send the releasable parking space identifier to the vehicle display screen so that the user can select a parking space.

[0158] It should be noted that the execution order of S308-S310 is not important.

[0159] In this embodiment, the implementation method and technical effect of S301-S314 are similar to those of the corresponding solutions in the above embodiments, and will not be repeated here.

[0160] Figure 4 This is a schematic diagram of the parking space management device provided in this application, as shown below. Figure 4 As shown, the parking space management device 40 provided in this embodiment includes: an acquisition module 41, a preliminary screening module 42, a verification module 43, and a sending module 44.

[0161] The system includes: an acquisition module 41 for acquiring environmental perception information of the target vehicle, including at least one candidate parking space; a preliminary screening module 42 for performing preliminary screening of each candidate parking space according to preset geometric rules to obtain a set of parking spaces to be verified; a verification module 43 for performing path feasibility verification based on the set of parking spaces to be verified using a preset path planning algorithm to generate a releasable parking space identifier; and a sending module 44 for sending the releasable parking space identifier to the vehicle display screen for users to select parking spaces.

[0162] The parking space management device provided in this embodiment can perform... Figure 2 and Figure 3 The implementation principles and technical effects of the methods shown are similar, and will not be repeated here.

[0163] Optionally, the preset geometric rules include a first set of rules and a second set of rules. The initial screening module 42, when performing preliminary screening of each candidate parking space information according to the preset geometric rules to obtain a set of parking spaces to be verified, is specifically used to: if all the rules in the first set of rules are satisfied, but no rule in the second set of rules is satisfied, then the candidate parking space information is stored in the set of parking spaces to be verified.

[0164] Optionally, the initial screening module 42, when performing preliminary screening of each candidate parking space information according to preset geometric rules to obtain a set of parking spaces to be verified, is specifically used for: if any rule in the first rule set is not met, then the candidate parking space information is stored in the non-release list; if all rules in the first rule set and the second rule set are met, then the candidate parking space information is stored in the release list.

[0165] Optionally, the set of parking spaces to be verified includes multiple parking space information to be verified. When the verification module 43 performs path feasibility verification based on the set of parking spaces to be verified using a preset path planning algorithm to generate a releasable parking space identifier, it is specifically used to: when any parking space information to be verified passes the path feasibility verification, generate a releasable parking space identifier based on the parking space information to be verified, and stop verifying the remaining parking space information to be verified.

[0166] Optionally, the parking space management device provided in this embodiment also includes an allocation module.

[0167] Correspondingly, the allocation module is used to distribute the path planning tasks of multiple parking spaces to be verified to multiple computing units for parallel execution, and synchronize the results through a shared memory mechanism.

[0168] Optionally, the parking space management device provided in this embodiment also includes a preprocessing module.

[0169] Correspondingly, the preprocessing module is used to eliminate parking spaces with preset obstacles in the parking area based on obstacle data in the environmental perception information; eliminate candidate parking spaces with perception confidence lower than a preset confidence threshold; and eliminate parking spaces with incomplete information such as missing corner points or abnormal parking space type identification.

[0170] Optionally, the parking space management device provided in this embodiment also includes a determination module.

[0171] Correspondingly, the determination module is used to determine the lateral distance threshold of the parking space entrance based on the obstacle density in the environmental perception information; and to determine the angle threshold between the parking space and the vehicle based on the parking space type.

[0172] Optionally, the verification module 43 is also used to perform dynamic verification and environmental real-time verification on the feasible path that has passed the feasibility verification. The dynamic verification includes determining whether the path meets the minimum turning radius of the vehicle, and the environmental real-time verification includes determining whether there are any newly appearing obstacles on the path.

[0173] Figure 5 A schematic diagram of the structure of the electronic device provided in this application. Figure 5 As shown, the electronic device 50 provided in this embodiment includes a processor 51 and a memory 52. ​​The processor 51 and the memory 52 are connected via a bus and communicate with each other.

[0174] In the specific implementation process, the processor 51 executes the computer execution instructions stored in the memory 52, causing the processor 51 to perform the above-described method.

[0175] The specific implementation process of processor 51 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0176] In the above embodiments, it should be understood that the processor 51 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0177] The memory 52 may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0178] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0179] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0180] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0181] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0182] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0183] The division of units is merely a logical functional division; 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 indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0184] 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.

[0185] In addition, the functional units in the various embodiments of the present invention 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.

[0186] If a function 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 invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0187] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0188] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A parking space management method, characterized in that, The method includes: Acquire environmental perception information of the target vehicle, wherein the environmental perception information includes at least one candidate parking space information; The information of each candidate parking space is initially screened according to preset geometric rules to obtain a set of parking spaces to be verified. A preset path planning algorithm is used to verify the feasibility of the path based on the set of parking spaces to be verified, so as to generate a releasable parking space identifier; The releasable parking space identifier is sent to the vehicle display screen so that the user can select a parking space.

2. The method according to claim 1, characterized in that, The preset geometric rules include a first set of rules and a second set of rules. The initial screening includes: If all rules in the first rule set are satisfied, but no rule in the second rule set is satisfied, then the candidate parking space information is stored in the parking space set to be verified.

3. The method according to claim 2, characterized in that, The initial screening includes: If any rule in the first set of rules is not met, the candidate parking space information is stored in the non-release list. If all rules in the first and second rule sets are satisfied, the candidate parking space information is stored in the release list.

4. The method according to claim 1, characterized in that, The set of parking spaces to be verified includes information on multiple parking spaces to be verified. Route feasibility verification includes: When any parking space information to be verified passes the path feasibility verification, a releaseable parking space identifier is generated based on that parking space information, and the verification of the remaining parking space information to be verified is stopped.

5. The method according to claim 4, characterized in that, Also includes: The path planning task for multiple parking spaces to be verified is distributed to multiple computing units for parallel execution, and the results are synchronized through a shared memory mechanism.

6. The method according to claim 1, characterized in that, Also includes: Based on obstacle data in environmental perception information, parking spaces containing preset obstacles are excluded from the parking space area. Eliminate candidate parking spaces whose perceived confidence level is lower than a preset confidence threshold; Remove parking spaces with incomplete information, such as those with missing corner points or abnormal parking space type identification.

7. The method according to claim 1, characterized in that, The preset geometric rules include the lateral distance threshold for parking space entrances, the angle threshold between the vehicle and the parking space, and also include: The lateral distance threshold for parking space entrances is determined based on the obstacle density in the environmental perception information. Determine the threshold for the angle between the parking space and the vehicle based on the type of parking space.

8. The method according to claim 1, characterized in that, Also includes: The feasible paths that have passed the feasibility verification are subjected to dynamic verification and environmental real-time verification. The dynamic verification includes determining whether the path meets the minimum turning radius of the vehicle, and the environmental real-time verification includes determining whether there are any new obstacles on the path.

9. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-8.

Images