Parking lot dispatching method, device, system and medium based on space-time decision

Abstract

The application relates to the technical field of vehicles, in particular to a parking lot scheduling method, device and system based on space-time decision and a medium, the method comprising the following steps: acquiring field end data and vehicle end data, aligning the field end data and the vehicle end data on a high-precision map at the field end, and integrating vehicle cooperative positioning data; performing real-time fusion on field end and vehicle end sensor data to obtain cooperative perception information of the whole road; integrating the real-time cooperative positioning of each vehicle, historical positioning data, the occupancy of each parking space and real-time cooperative perception data of each vehicle and the field end in time and space to obtain complete traffic flow in the field; constructing a target function of optimal scheduling based on traffic states and perception states, and solving the target function to obtain optimal scheduling results of the traffic flow; the application makes full use of actual data of the field end and the vehicle end, can improve parking space utilization, reduce vehicle congestion and optimize the parking experience.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, and specifically to a parking lot scheduling method, device, system, and medium based on spatiotemporal decision-making. Background Technology

[0002] Parking lot vehicle planning and scheduling is a crucial component of smart city construction. With accelerated urbanization, increasing parking demand, and rapid technological advancements, parking lot scheduling and management have evolved from traditional manual methods towards intelligent and information-based approaches. However, current parking lot vehicle planning and scheduling still face many challenges and shortcomings.

[0003] Currently, parking lots are still dominated by traditional parking lots. Among these traditional parking lots, most small ones still rely on manual payment and management, which is inefficient and prone to human error. Secondly, traditional parking lots cannot obtain real-time vehicle entry and exit information, resulting in low parking space utilization. Thirdly, traditional parking space allocation is usually fixed, without dynamic scheduling and optimization, making it difficult to maximize the use of parking lot resources.

[0004] With the rise of smart parking lots, some have introduced technologies such as license plate recognition, automatic payment (ETC / contactless payment), and electronic fences, reducing manual intervention. Furthermore, some parking lots in certain cities have achieved network connectivity, allowing users to check parking availability via apps or platforms. Some smart parking lots support online parking space reservations and navigation, but their adoption rate is limited. Technically, due to the lack of unified standards and interfaces in independent development, data silos often exist between parking lots and between parking lots and vehicles, making data sharing and collaborative scheduling difficult. In addition, most existing scheduling algorithms are based on simple rule guidance and do not fully utilize real-time data and predictive models for dynamic optimization. In complex environments (such as multi-story or multi-traffic intersection parking lots), the scheduling system performs inadequately. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a parking lot scheduling method, device, system and medium based on spatiotemporal decision-making, which uses vehicle-road-cloud integration technology to avoid data silos between parking lots and between parking lots and vehicles, and uses spatiotemporal decision-making algorithms to make full use of the actual data at the parking lot end and the vehicle end, thereby improving the current problem of simple parking lot guidance and poor handling of complex environments.

[0006] On one hand, embodiments of the present invention provide a parking lot scheduling method based on spatiotemporal decision-making, the method comprising the following steps: Acquire site-side data and vehicle-side data, align the site-side data and vehicle-side data on the high-precision map at the site, and integrate them to obtain vehicle collaborative positioning data; Real-time fusion of data from field and vehicle sensors yields collaborative perception information for the entire road area; The real-time collaborative positioning of each vehicle, historical positioning data, parking space occupancy status, and real-time collaborative perception data between each vehicle and the site are integrated in time and space to obtain the complete traffic flow in the site. The objective function for optimal scheduling is constructed based on traffic conditions and perception conditions. Solving the objective function yields the optimal scheduling result for traffic flow.

[0007] Optionally, aligning the field data and vehicle data on a high-precision map at the field and integrating them to obtain vehicle cooperative positioning data includes: Time synchronization is performed between field-end data and vehicle-end data. The field-end data includes field-end positioning data, field-end images, and field-end radar data, while the vehicle-end data includes vehicle-end positioning data, vehicle-end images, IMU data, and vehicle-end radar data. Using the high-precision map at the field terminal as a unified benchmark, the field terminal positioning data and vehicle terminal positioning data are aligned to the high-precision map; geometric feature matching based on point cloud is used to register the field terminal radar data and vehicle terminal radar data; and visual SLAM based on image feature points is used to align the field terminal image and vehicle terminal image to the high-precision map. Based on the diffusion Kalman filter, vehicle-side data and field-side data are integrated, and global collaborative localization is performed on the integrated data based on graph optimization to obtain vehicle collaborative localization data.

[0008] Optionally, the real-time fusion of data from field-end and vehicle-end sensors to obtain collaborative perception information for the entire road area includes: Edge computing devices are used to preprocess field-end image data and field-end radar data to obtain multiple field-end sensing results; The vehicle-side perception results are generated from IMU data and radar data of the surrounding environment. A confidence-weighted fusion method is used to assign weights to the field perception results and the vehicle perception results, and the fused data is obtained based on the weighted results. Redundant and conflicting fused data is eliminated through spatiotemporal consistency checks, and globally consistent collaborative perception data is generated using multi-target tracking.

[0009] Optionally, the real-time collaborative positioning of each vehicle, historical positioning data, parking space occupancy status, and real-time collaborative sensing data between each vehicle and the site are integrated temporally and spatially to obtain a complete traffic flow within the site, including: Acquire real-time data and historical data. The real-time data includes vehicle cooperative positioning data, cooperative sensing data, and parking space occupancy information. The historical data includes historical vehicle positioning data, historical parking space occupancy data, and historical data of traffic participants. The traffic status at the current time is obtained by weighted fusion of real-time data and historical data within the time window; The collaborative perception data of the entire road area is weighted and fused with confidence level to obtain the perception status at the current time.

[0010] Optionally, the weighted fusion of real-time data and historical data within the time window to obtain the traffic status at the current time includes: Align the vehicle-end positioning data and site-end positioning data in the vehicle cooperative positioning data to the map coordinate system of the high-precision map to obtain the vehicle-end position and site-end position; perform confidence weighting on the vehicle-end position and site-end position to obtain the unified position of spatial fusion. Align the historical vehicle location data and historical field location data from the vehicle historical location data within the most recent time window to the map coordinate system of the high-precision map to obtain the historical vehicle location and historical field location; obtain the unified location fused by time based on the historical vehicle location and historical field location; By performing spatiotemporal integration and weighting on the unified location of spatial fusion and the unified location of temporal fusion, the traffic status at the current time is obtained.

[0011] Optionally, obtaining the unified location based on historical vehicle-end location and historical field-end location through time fusion includes: The historical vehicle-end position and historical field-end position at each time point within the most recent time window are weighted by confidence to obtain the unified position at each time point. The average of the unified locations at each time point is calculated to obtain the unified location of time fusion.

[0012] Optionally, the step of constructing an objective function for optimal scheduling based on traffic conditions and perception conditions, and solving the objective function to obtain the optimal scheduling result of traffic flow, includes: Extract traffic flow features from traffic conditions and perception flow features from perception conditions, respectively; Global feature aggregation is obtained based on context-aware traffic flow features, and a road segment delay correction term is generated based on the global traffic features. Traffic flow distribution is calculated based on the traffic characteristics between nodes in the road segment. Capacity load is generated based on the traffic flow distribution and the capacity of the corresponding road segment. A preliminary objective function for optimal scheduling is constructed based on the capacity load and the correction term for road segment delay. Based on the preliminary objective function and the characteristics of the perceived flow, an objective function for optimal scheduling is constructed, and the optimal scheduling result of the traffic flow is obtained by solving the objective function.

[0013] On the other hand, embodiments of the present invention provide a parking lot scheduling device based on spatiotemporal decision-making, comprising: The collaborative positioning module is used to acquire site data and vehicle data, align the site data and vehicle data on the high-precision map at the site, and integrate them to obtain vehicle collaborative positioning data; The collaborative perception module is used to fuse data from field and vehicle sensors in real time to obtain collaborative perception information for the entire road area. The spatiotemporal integration module is used to integrate the real-time collaborative positioning of each vehicle and historical positioning data, the occupancy status of each parking space, and the real-time collaborative perception data between each vehicle and the site to obtain the complete traffic flow in the site. The collaborative scheduling module is used to construct the objective function for optimal scheduling based on traffic conditions and perception conditions, and to solve the objective function to obtain the optimal scheduling result of traffic flow.

[0014] On the other hand, embodiments of the present invention provide a parking lot scheduling system based on spatiotemporal decision-making, comprising: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor performs the method described above.

[0015] On the other hand, embodiments of the present invention provide a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to perform the above-described method.

[0016] The embodiments of the present invention have the following beneficial effects: This invention proposes a parking lot scheduling method, device, system, and medium based on spatiotemporal decision-making. It acquires parking lot-side and vehicle-side data, aligns these data on a high-precision map at the parking lot, and integrates them to obtain vehicle cooperative positioning data. Then, it fuses parking lot-side and vehicle-side sensor data in real time to obtain cooperative perception information across the entire road area. This information is then combined with real-time and historical positioning data of each vehicle and parking space occupancy status for spatiotemporal integration. Finally, it constructs and solves the optimal scheduling objective function to achieve efficient parking lot traffic flow scheduling. This invention utilizes vehicle-road-cloud integration technology to avoid data silos between parking lots and between parking lots and vehicles. It uses a spatiotemporal decision-making algorithm to fully utilize actual data from both the parking lot and vehicle sides, improving upon the current problem of simple parking lot guidance and poor handling of complex environments. Through precise scheduling, this invention improves parking space utilization, reduces traffic congestion, and optimizes the parking experience. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the steps of a parking lot scheduling method based on spatiotemporal decision-making provided in an embodiment of the present invention. Figure 2This is an architecture diagram of a parking lot scheduling scheme based on spatiotemporal decision-making provided by an embodiment of the present invention; Figure 3 This is a development architecture diagram of the data acquisition module provided in an embodiment of the present invention; Figure 4 This is a structural block diagram of a parking lot scheduling device based on spatiotemporal decision-making provided in an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0019] It should be noted that although the device diagram shows a modular division and the flowchart illustrates a logical order, in some cases, the steps shown or described may be performed in a different order than the modular division in the device or the order shown in the flowchart. The terms "first," "second," etc., used in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0021] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a thorough understanding of embodiments of this application. However, those skilled in the art will recognize that the technical solutions of this application can be practiced without one or more of the specific details, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of this application.

[0022] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0023] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0024] like Figure 1 and Figure 2 As shown, Figure 1 A parking lot scheduling method based on spatiotemporal decision-making is provided in this embodiment of the invention. The method includes the following steps: S100 acquires site data and vehicle data, aligns the site data and vehicle data on the high-precision map at the site, and integrates them to obtain vehicle cooperative positioning data; Specifically, the field-end data includes field-end positioning data, image data, and field-end radar data, while the vehicle-end data includes vehicle-end positioning data, IMU data, and vehicle-end radar data. The field-end and vehicle-end positioning data are fused in real-time in the cloud and synchronously interfaced with the Global Positioning System (GPS) coordinate system (WGS84), the Chinese coordinate system (GCJ-02), and the Baidu coordinate system (BD-09) to achieve a universal collaborative positioning interface. Simultaneously, the real-time vehicle location is accurately projected onto the high-precision map at the field-end, thereby achieving centimeter-level precise vehicle positioning.

[0025] The S200 integrates data from field and vehicle sensors in real time to obtain collaborative perception information for the entire road area. Specifically, the system integrates data from both the parking lot and vehicle sensors in real time in the cloud, allowing them to complement and verify each other to achieve real-time collaborative perception between the parking lot and the vehicle. Based on the perception information obtained across the entire road area, it provides a data foundation for more efficient scheduling of each vehicle in the parking lot.

[0026] The S300 integrates real-time collaborative positioning of each vehicle with historical positioning data, occupancy status of each parking space, and real-time collaborative perception data between each vehicle and the site to obtain a complete traffic flow in the site. Specifically, the real-time collaborative positioning of each vehicle in the parking lot, as well as historical positioning data, the occupancy status of each parking space, and the real-time collaborative perception data between each vehicle and the parking lot are integrated in the cloud in terms of time (historical data and real-time data) and space (multi-layer positioning and traffic participation) to obtain a complete traffic flow situation in the parking lot as the output of the scheduling module.

[0027] S400 constructs an objective function for optimal scheduling based on traffic conditions and perception conditions, and solves the objective function to obtain the optimal scheduling result of traffic flow.

[0028] Specifically, based on the spatiotemporal fusion results, the transformer can extract the correlation between the input data before and after the input data, and can extract the associated features in the streaming state of the integrated output of localization and perception, thereby performing global scheduling based on the extracted features.

[0029] Steps S100 to S400, as illustrated in this embodiment, firstly, by acquiring site data and vehicle data, the site data and vehicle data are aligned on a high-precision map at the site to obtain vehicle cooperative positioning data. Then, by real-time fusion of site and vehicle sensor data, cooperative perception information for the entire road area is obtained. This is then combined with real-time and historical positioning data of each vehicle and parking space occupancy status for spatiotemporal integration. Finally, the optimal scheduling objective function is constructed and solved to achieve efficient scheduling of parking lot traffic flow. This invention utilizes vehicle-road-cloud integrated technology to avoid data silos between parking lots and between parking lots and vehicles. It uses a spatiotemporal decision algorithm to fully utilize actual data from both the site and vehicle sides, improving upon the current problem of simple parking lot guidance and poor handling of complex environments.

[0030] In some embodiments, aligning the field data and vehicle data on a high-precision map at the field and integrating them to obtain vehicle cooperative positioning data includes: Time synchronization is performed between field-end data and vehicle-end data. The field-end data includes field-end positioning data, field-end images, and field-end radar data, while the vehicle-end data includes vehicle-end positioning data, vehicle-end images, IMU data, and vehicle-end radar data. Using the high-precision map at the field terminal as a unified benchmark, the field terminal positioning data and vehicle terminal positioning data are aligned to the high-precision map; geometric feature matching based on point cloud is used to register the field terminal radar data and vehicle terminal radar data; and visual SLAM based on image feature points is used to align the field terminal image and vehicle terminal image to the high-precision map. Based on the diffusion Kalman filter, vehicle-side data and field-side data are integrated, and global collaborative localization is performed on the integrated data based on graph optimization to obtain vehicle collaborative localization data.

[0031] The specific steps are as follows: Time synchronization of multi-source data: Based on GPS timestamps or PTP protocol, the positioning data of field and vehicle devices are accurately synchronized to ensure data fusion consistency.

[0032] Data registration and high-precision map alignment: Using the high-precision map at the field as a unified benchmark, the positioning data from the vehicle and the field are aligned. Geometric feature matching based on point clouds is used to register the LiDAR point cloud data, and finally, visual SLAM based on image feature points is used to align it with the map.

[0033] Data fusion: Integrating vehicle-side and field-side data based on diffused Kalman filtering. Key information includes GNSS data (providing global positioning), IMU data (providing local motion trajectories), and field-side sensor data (providing vehicle position constraint information).

[0034] Global collaborative localization based on graph optimization.

[0035] In some embodiments, the real-time fusion of field-end and vehicle-end sensor data to obtain collaborative perception information for the entire road area includes: Edge computing devices are used to preprocess field-end image data and field-end radar data to obtain multiple field-end sensing results; The vehicle-side perception results are generated from IMU data and radar data of the surrounding environment. A confidence-weighted fusion method is used to assign weights to the field perception results and the vehicle perception results, and the fused data is obtained based on the weighted results. Redundant and conflicting fused data is eliminated through spatiotemporal consistency checks, and globally consistent collaborative perception data is generated using multi-target tracking.

[0036] The specific plan is as follows: Data access includes both site-side perception and vehicle-side perception. For site-side perception, road cameras and LiDAR are responsible for detecting traffic participants (such as vehicles, pedestrians, and non-motorized vehicles). Image data is collected by road cameras, and radar data is collected by LiDAR. Edge computing devices are used to preprocess the data. For vehicle-side perception, vehicle sensors are responsible for perceiving the surrounding environment and generating preliminary results using the YOLO algorithm running locally on the vehicle.

[0037] Multimodal data fusion: A confidence-weighted fusion method is used to assign weights to the detection results of each sensor, and fusion is performed based on the weighted results. Then, a spatiotemporal consistency check is performed to eliminate redundant and conflicting perception results. Finally, multi-target tracking is used to generate globally consistent target trajectories.

[0038] In some embodiments, the real-time collaborative positioning of each vehicle, historical positioning data, parking space occupancy status, and real-time collaborative sensing data between each vehicle and the site are integrated temporally and spatially to obtain a complete traffic flow in the site, including: Acquire real-time data and historical data. The real-time data includes vehicle cooperative positioning data, cooperative sensing data, and parking space occupancy information. The historical data includes historical vehicle positioning data, historical parking space occupancy data, and historical data of traffic participants. The traffic status at the current time is obtained by weighted fusion of real-time data and historical data within the time window; The collaborative perception data of the entire road area is weighted and fused with confidence level to obtain the perception status at the current time.

[0039] The specific plan is as follows: Data input: Real-time data sources include vehicle cooperative positioning data, real-time cooperative sensing data, and parking space occupancy information; Vehicle cooperative positioning data: High-precision vehicle location generated by fusing vehicle-side sensors (GNSS, IMU, LiDAR, etc.) and field-side equipment (RSU, road cameras, LiDAR, etc.).

[0040] Real-time collaborative perception data: including vehicle detection, pedestrian tracking, obstacle recognition, etc.

[0041] Parking space occupancy information: Updated in real time via parking lot sensors (such as geomagnetic sensors, cameras, and Bluetooth beacons).

[0042] Historical data sources include historical vehicle location data, historical parking space occupancy data, and historical traffic participant data; Historical vehicle location data: records the vehicle's trajectory for traffic flow analysis.

[0043] Historical parking space occupancy data: used to analyze parking lot usage patterns.

[0044] Historical data of traffic participants: historical trajectories of pedestrians and non-motorized vehicles.

[0045] Data processing: Assume the real-time data stream is... , This represents the data collected at time t (such as vehicle location, speed, and location of other road users). Combining this data, the traffic state at the current time t is calculated. .

[0046] The sliding window method assumes that historical data within a time window T has a significant impact on the current state. Considering the different importance of real-time and historical data, a weighted fusion of historical and real-time data can be performed, using the following formula:

[0047] in: It refers to the traffic status at time point t. It is real-time data at time point t. It represents the real-time data at time point i, and T is the length of the sliding window (e.g., data from the past 10 minutes). These are weighting coefficients, representing the importance of real-time data (usually...). The larger the value, the higher the weight of the real-time data. This represents the average value of historical data within a past window T.

[0048] For multi-source sensing data (such as vehicle-mounted cameras and field radar), target detection results can be fused using confidence-weighted fusion, as shown in the formula: ;in: It is the sensing result of the i-th sensor. It represents the confidence level of the sensor.

[0049] In some embodiments, the weighted fusion of real-time data and historical data within a time window to obtain the traffic status at the current time includes: Align the vehicle-end positioning data and site-end positioning data in the vehicle cooperative positioning data to the map coordinate system of the high-precision map to obtain the vehicle-end position and site-end position; perform confidence weighting on the vehicle-end position and site-end position to obtain the unified position of spatial fusion. Align the historical vehicle location data and historical field location data from the vehicle historical location data within the most recent time window to the map coordinate system of the high-precision map to obtain the historical vehicle location and historical field location; obtain the unified location fused by time based on the historical vehicle location and historical field location; By performing spatiotemporal integration and weighting on the unified location of spatial fusion and the unified location of temporal fusion, the traffic status at the current time is obtained.

[0050] Specifically, let the vehicle-side positioning data be... Field-end positioning data is High-precision maps are We need to and Align to high-precision map Under a unified coordinate system, the positions of traffic participants are obtained uniformly. The core of spatial integration is coordinate transformation and multi-source data fusion. Assuming that the data from the vehicle and the field are in their respective coordinate systems, they can be aligned using a coordinate transformation matrix. Assuming real-time point cloud data is Map point cloud data is Then we can find a transformation matrix B such that:

[0051] Where B is the transformation matrix. , These are the real-time point cloud data and the map point cloud data at time point t, respectively.

[0052] Therefore, there is a transformation matrix to convert vehicle-side and field-side data to the map coordinate system. and Then we have:

[0053] in: It refers to the vehicle's location in the map coordinate system, which corresponds to the vehicle's location data. It is the field location corresponding to the field location data in the map coordinate system.

[0054] The final unified position is:

[0055] in: and It is the confidence weight of vehicle-side and site-side data (for example, site-side data may have higher accuracy due to fixed installation equipment). It is a unified position. The size is N is the number of positioning points, and d is the coordinate dimension.

[0056] In some embodiments, obtaining the unified location based on historical vehicle-end location and historical field-end location through time fusion includes: The historical vehicle-end position and historical field-end position at each time point within the most recent time window are weighted by confidence to obtain the unified position at each time point. The average of the unified locations at each time point is calculated to obtain the unified location of time fusion.

[0057] Specifically, based on the sliding window, the final positioning state integration formula can be expressed as a combination of time and space:

[0058] in: This represents the traffic state at time point t after spatiotemporal integration. The first term in the formula is the real-time state (spatial fusion), and the second term is the temporal fusion of the historical state. The size is N is the number of positioning points, and d is the coordinate dimension.

[0059] In some embodiments, constructing an objective function for optimal scheduling based on traffic conditions and perception conditions, and solving the objective function to obtain the optimal scheduling result of the traffic flow, includes: Extract traffic flow features from traffic conditions and perception flow features from perception conditions, respectively; Global feature aggregation is obtained based on context-aware traffic flow features, and a road segment delay correction term is generated based on the global traffic features. Traffic flow distribution is calculated based on the traffic characteristics between nodes in the road segment. Capacity load is generated based on the traffic flow distribution and the capacity of the corresponding road segment. A preliminary objective function for optimal scheduling is constructed based on the capacity load and the correction term for road segment delay. Based on the preliminary objective function and the characteristics of the perceived flow, an objective function for optimal scheduling is constructed, and the optimal scheduling result of the traffic flow is obtained by solving the objective function.

[0060] Specifically, refer to Figure 3 Let Transformer1 be... Transformer2 is Traffic flow characteristics are The characteristics of the perceptual flow are Therefore, we have: , ; Indicates vehicle coordinates. This represents the vehicle coordinates after spatiotemporal integration at time t. Traffic conditions at the location.

[0061] Suppose context-aware traffic flow features Therefore, there is global feature aggregation:

[0062] in, This indicates global feature aggregation. The total number T represents the total number of time windows. Indicates the number of the location point. Indicates the time window number, The first characteristic in traffic flow is represented by the second characteristic. Traffic characteristics of the i-th location within a given time window. It is a d-dimensional vector.

[0063] When only traffic flow is considered, the objective of scheduling algorithm optimization is:

[0064] in, This represents the traffic flow distribution calculated using the features of any two nodes. This indicates the capacity of the corresponding road segment. This represents the traffic flow characteristics at the i-th location point. This represents the traffic flow characteristics at the j-th location point. This represents the range of values ​​for i and j. Corresponding to [1, N], .

[0065] It is a correction term for road segment delay extracted by global traffic features from the Transformer, minimizing... This allows us to obtain the optimal scheduling considering only traffic flow. However, in actual operation, real-time sensing information also affects the scheduling, hence:

[0066] This represents the sensing results between each node from the current time to the next time, so it is minimized. This allows us to obtain the optimal scheduling that only considers traffic flow.

[0067] refer to Figure 4 The diagram shown illustrates the architecture of a parking lot scheduling device based on spatiotemporal decision-making, which includes a collaborative positioning module, a collaborative sensing module, a spatiotemporal integration module, and a collaborative scheduling module.

[0068] The collaborative positioning module is used to acquire site data and vehicle data, align the site data and vehicle data on the high-precision map at the site, and integrate them to obtain vehicle collaborative positioning data; The collaborative positioning module integrates positioning data from the field and the vehicle in real time in the cloud and synchronously interfaces with the Global Positioning System coordinate system WGS84, the Chinese coordinate system GCJ-02, and the Baidu coordinate system BD-09 to achieve universal collaborative positioning interface. At the same time, it accurately projects the real-time position of the vehicle onto the high-precision map at the field, thereby achieving centimeter-level accurate positioning of the vehicle.

[0069] The collaborative perception module is used to fuse data from field and vehicle sensors in real time to obtain collaborative perception information for the entire road area. The collaborative perception module integrates data from both on-site and vehicle sensors in real time in the cloud, complementing and verifying each other to achieve real-time collaborative perception between vehicles and parking lots. Based on the perception information obtained across the entire road area, it provides a data foundation for more efficient scheduling of each vehicle in the parking lot.

[0070] The spatiotemporal integration module is used to integrate the real-time collaborative positioning of each vehicle and historical positioning data, the occupancy status of each parking space, and the real-time collaborative perception data between each vehicle and the site to obtain the complete traffic flow in the site. The spatiotemporal integration module integrates real-time collaborative positioning and historical positioning data of each vehicle in the parking lot, the occupancy status of each parking space, and real-time collaborative perception data between each vehicle and the parking lot in the cloud, in terms of time (historical data and real-time data) and space (multi-layer positioning and traffic participation), to obtain a complete traffic flow situation in the parking lot as the output of the scheduling module.

[0071] The collaborative scheduling module is used to construct the objective function for optimal scheduling based on traffic conditions and perception conditions, and to solve the objective function to obtain the optimal scheduling result of traffic flow.

[0072] Specifically, based on the spatiotemporal fusion results, the transformer can extract the correlation between the input data before and after the input data, and can extract the associated features in the streaming state of the integrated output of localization and perception, thereby performing global scheduling based on the extracted features.

[0073] This invention also provides a parking lot scheduling system based on spatiotemporal decision-making, including a memory, a processor, and a program stored in the memory and executable on the processor. When the program is executed by the processor, it implements the method described in the above embodiments.

[0074] Taking the example of a processor and memory in a vehicle controller being connected via a bus, the memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory may optionally include memory remotely located relative to the control processor, and these remote memories can be connected to the control device via a network.

[0075] The non-transitory software program and instructions required to implement the methods of the above embodiments are stored in memory and executed by the processor to perform the methods of the above embodiments.

[0076] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0077] This invention proposes a parking lot scheduling method, device, system, and medium based on spatiotemporal decision-making. Compared with related technologies, this invention has the following advantages: 1. By comprehensively utilizing the characteristics of time and space information, we can achieve full perception of information on all dispatched vehicles in parking lot scenarios.

[0078] 2. Use sensory integration information to assist in the scheduling of traffic flow across the entire region and optimize the scheduling logic.

[0079] This invention also provides a vehicle including the control device described in the above embodiments.

[0080] The vehicle can be a private car, such as a sedan, SUV, MPV, or pickup truck. It can also be a commercial vehicle, such as a van, bus, small truck, or large semi-trailer. The vehicle must have an electric motor capable of outputting power or acting as a generator to store mechanical energy. When the vehicle is a new energy vehicle, it can be a hybrid or a pure electric vehicle.

[0081] Since the vehicle applies all the technical solutions of the above-mentioned control device or vehicle controller, it has at least all the beneficial effects brought about by the technical solutions of the above embodiments, which will not be repeated here.

[0082] Furthermore, one embodiment of the present invention provides a computer-readable storage medium storing computer-executable instructions for performing the above-described method.

[0083] It is worth noting that, since the computer-readable storage medium of the present invention is capable of executing the methods of any of the above embodiments, the specific implementation methods and technical effects of the computer-readable storage medium of the present invention can be referred to the specific implementation methods and technical effects of the methods of any of the above embodiments.

[0084] Furthermore, one embodiment of the present invention provides a computer program product, including a computer program or computer instructions, the computer program or computer instructions being stored in a computer-readable storage medium, a processor of a computer device reading the computer program or computer instructions from the computer-readable storage medium, and the processor executing the computer program or computer instructions to cause the computer device to perform the above-described method.

[0085] It is worth noting that, since the computer program product of the present invention can execute the methods of any of the above embodiments, the specific implementation methods and technical effects of the computer program product of the present invention can be referred to the specific implementation methods and technical effects of the methods of any of the above embodiments.

[0086] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically include computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0087] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

Claims

1. A parking lot scheduling method based on spatiotemporal decision-making, characterized in that, The method includes: Acquire site-side data and vehicle-side data, align the site-side data and vehicle-side data on the high-precision map at the site, and integrate them to obtain vehicle collaborative positioning data; Real-time fusion of data from field and vehicle sensors yields collaborative perception information for the entire road area; The system integrates real-time and historical location data of each vehicle, parking space occupancy status, and real-time collaborative perception data between each vehicle and the site to obtain a complete traffic flow within the site. Specifically, this includes: acquiring real-time and historical data, whereby real-time data includes vehicle collaborative positioning data, collaborative perception data, and parking space occupancy information; and historical data includes historical vehicle positioning data, historical parking space occupancy data, and historical data of traffic participants. The system then performs weighted fusion of real-time data and historical data within a time window to obtain the current traffic status; and performs confidence-weighted fusion of collaborative perception data across the entire road area to obtain the current perception status. The optimal scheduling objective function is constructed based on traffic state and perceived state, and the optimal scheduling result of traffic flow is obtained by solving the objective function. Specifically, this includes: extracting traffic flow features from traffic state and perceived flow features from perceived state; obtaining global feature aggregation based on context-aware traffic flow features, generating a road segment delay correction term based on global traffic features; calculating traffic flow distribution based on traffic features between nodes in the road segment, generating capacity load based on traffic flow distribution and corresponding road segment capacity, constructing a preliminary objective function for optimal scheduling based on capacity load and road segment delay correction term; constructing the optimal scheduling objective function based on the preliminary objective function and perceived flow features, and solving the objective function to obtain the optimal scheduling result of traffic flow.

2. The method according to claim 1, characterized in that, The process of aligning and integrating field-side data and vehicle-side data on a high-precision map at the field site to obtain vehicle cooperative positioning data includes: Time synchronization is performed between field-end data and vehicle-end data. The field-end data includes field-end positioning data, field-end images, and field-end radar data, while the vehicle-end data includes vehicle-end positioning data, vehicle-end images, IMU data, and vehicle-end radar data. Using the high-precision map at the field terminal as a unified benchmark, the field terminal positioning data and vehicle terminal positioning data are aligned to the high-precision map; geometric feature matching based on point cloud is used to register the field terminal radar data and vehicle terminal radar data; and visual SLAM based on image feature points is used to align the field terminal image and vehicle terminal image to the high-precision map. Based on the diffusion Kalman filter, vehicle-side data and field-side data are integrated, and global collaborative localization is performed on the integrated data based on graph optimization to obtain vehicle collaborative localization data.

3. The method according to claim 1, characterized in that, The real-time fusion of data from field-end and vehicle-end sensors to obtain collaborative perception information for the entire road area includes: Edge computing devices are used to preprocess field-end image data and field-end radar data to obtain multiple field-end sensing results; The vehicle-side perception results are generated from IMU data and radar data of the surrounding environment. A confidence-weighted fusion method is used to assign weights to the field perception results and the vehicle perception results, and the fused data is obtained based on the weighted results. Redundant and conflicting fused data is eliminated through spatiotemporal consistency checks, and globally consistent collaborative perception data is generated using multi-target tracking.

4. The method according to claim 1, characterized in that, The weighted fusion of real-time data and historical data within the time window to obtain the traffic status at the current time includes: Align the vehicle-end positioning data and site-end positioning data in the vehicle cooperative positioning data to the map coordinate system of the high-precision map to obtain the vehicle-end position and site-end position; perform confidence weighting on the vehicle-end position and site-end position to obtain the unified position of spatial fusion. Align the historical vehicle location data and historical field location data from the vehicle historical location data within the most recent time window to the map coordinate system of the high-precision map to obtain the historical vehicle location and historical field location; obtain the unified location fused by time based on the historical vehicle location and historical field location; By performing spatiotemporal integration and weighting on the unified location of spatial fusion and the unified location of temporal fusion, the traffic status at the current time is obtained.

5. The method according to claim 4, characterized in that, The unified location obtained by time fusion based on historical vehicle-end location and historical field-end location includes: The historical vehicle-end position and historical field-end position at each time point within the most recent time window are weighted by confidence to obtain the unified position at each time point; The average of the unified locations at each time point is calculated to obtain the unified location of time fusion.

6. A parking lot scheduling device based on spatiotemporal decision-making, characterized in that, The device includes: The collaborative positioning module is used to acquire site data and vehicle data, align the site data and vehicle data on the high-precision map at the site, and integrate them to obtain vehicle collaborative positioning data; The collaborative perception module is used to fuse data from field and vehicle sensors in real time to obtain collaborative perception information for the entire road area. The spatiotemporal integration module is used to integrate real-time collaborative positioning and historical positioning data of each vehicle, parking space occupancy status, and real-time collaborative perception data between each vehicle and the site to obtain a complete traffic flow in the site. Specifically, it includes: acquiring real-time data and historical data, wherein the real-time data includes vehicle collaborative positioning data, collaborative perception data, and parking space occupancy information; the historical data includes historical vehicle positioning data, historical parking space occupancy data, and historical data of traffic participants; weightedly fusing the real-time data and historical data within the time window to obtain the traffic status at the current time; and performing confidence-weighted fusing of the collaborative perception data across the entire road area to obtain the perception status at the current time. The collaborative scheduling module is used to construct an objective function for optimal scheduling based on traffic conditions and perception conditions, and solve the objective function to obtain the optimal scheduling result for traffic flow. Specifically, it includes: extracting traffic flow features from traffic conditions and perception flow features from perception conditions; obtaining global feature aggregation based on context-aware traffic flow features, and generating a road segment delay correction term based on global traffic features; calculating traffic flow distribution based on traffic features between nodes in the road segment, generating capacity load based on traffic flow distribution and the capacity of the corresponding road segment, constructing a preliminary objective function for optimal scheduling based on capacity load and road segment delay correction term; constructing the optimal scheduling objective function based on the preliminary objective function and perception flow features, and solving the objective function to obtain the optimal scheduling result for traffic flow.

7. A parking lot scheduling system based on spatiotemporal decision-making, characterized in that, It includes a memory, a processor, and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the method of any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 5.

Images