A city-level parking resource optimization method and system

Abstract

The application discloses a kind of urban-level parking resource optimization method and system, it is related to intelligent transportation technical field.The method comprises: obtaining the real-time operation data and geographic position of target city area parking lot;The city parking topology graph is constructed with each parking lot as node, with connecting road as edge;Based on real-time traffic data, the real-time weight of each edge is obtained;Real-time parking pressure value is calculated;The node of real-time parking pressure value greater than or equal to first preset pressure value is high pressure node, and the node less than first preset pressure value is determined as potential node;Pressure conduction simulation is carried out, and simulation parking pressure value is obtained;The potential node and high pressure node with simulation parking pressure value greater than second preset pressure value and less than first preset pressure value are determined as potential congestion cluster;For each potential congestion cluster: based on simulation parking pressure value, calculate the pricing coefficient, obtain real-time parking rate, generate flow direction instruction.The application can improve the configuration accuracy of urban-level parking resource.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation technology, and in particular to a method and system for optimizing city-level parking resources. Background Technology

[0002] With the acceleration of urbanization and the continuous growth of motor vehicle ownership, the imbalance between supply and demand of urban parking resources in time and space has become increasingly prominent. To improve management efficiency, existing technologies have shifted from static information dissemination to proactive guidance based on prediction, optimizing traffic flow by anticipating future parking space availability.

[0003] Existing technologies employ spatiotemporal prediction models, such as combining Graph Convolutional Networks (GCNs) and Long Short-Term Memory Networks (LSTMs), to model urban parking networks as graph structures. GCNs capture the geographical proximity relationships between parking lots, while LSTMs learn the temporal evolution patterns of historical parking occupancy rates. By identifying recurring patterns in historical data, relatively accurate predictions can be made about changes in parking spaces under normal conditions.

[0004] However, existing predictive models have limitations in identifying the spatial transmission effects of parking pressure. When parking lots in a certain area reach saturation, the overflowing parking demand spreads to surrounding areas, creating a chain reaction. While existing GCN-based models can capture geographical proximity, they cannot dynamically reflect the impact of real-time traffic conditions on pressure transmission, leading to inaccurate predictions of parking pressure in surrounding areas. This affects the accuracy of pricing strategies and their effectiveness in attracting traffic, ultimately reducing the accuracy of city-level parking resource allocation. Summary of the Invention

[0005] This application provides a method and system for optimizing city-level parking resources, which addresses the technical problem of improving the accuracy of city-level parking resource allocation.

[0006] The technical solution of this application embodiment is implemented as follows: In a first aspect, embodiments of this application provide a city-level parking resource optimization method, including: Obtain real-time operational data and geographic location of all parking lots within the target city area; Construct an urban parking topology graph with each parking lot as a node and the connecting roads of each node as edges; Based on the real-time traffic data of the target city area, the real-time weights of each edge are obtained; Based on the real-time operational data, the real-time parking pressure value of each node is calculated, wherein the real-time operational data includes at least the static capacity and real-time occupancy rate of the corresponding parking lot; Nodes whose real-time parking pressure value is greater than or equal to the first preset pressure value are identified as high-pressure nodes, and nodes whose real-time parking pressure value is less than the first preset pressure value are identified as potential nodes. Based on the real-time weights, pressure transmission simulation is performed on each high-pressure node in the urban parking topology map to obtain the simulated parking pressure value of each node. The potential nodes whose simulated parking pressure value is greater than the second preset pressure value and less than the first preset pressure value, together with a corresponding high-pressure node, are identified as a potential congestion cluster. The second preset pressure value is less than the first preset pressure value, and the actual traffic distance between the potential node and the corresponding high-pressure node is less than a preset distance threshold. For each of the aforementioned potential congested clusters, the following steps are performed to optimize resources: Based on the simulated parking pressure value corresponding to each target node, calculate the price adjustment coefficient corresponding to each target node, where the target node is any node of the potential congestion cluster; Based on the price adjustment coefficient, the real-time parking fee rate after price adjustment for each target node is obtained; Based on the real-time parking rates mentioned above, a referral instruction is generated.

[0007] Optionally, obtaining the real-time weight of each edge based on the real-time traffic data of the target city area includes: obtaining the actual traffic distance and preset traffic speed corresponding to each edge; calculating the ratio of each actual traffic distance to the preset traffic speed to obtain the simulated travel time corresponding to each edge; and calculating the reciprocal of each simulated travel time to obtain the real-time weight corresponding to each edge.

[0008] Optionally, the real-time operation data also includes the entry count and exit count of the corresponding node within a preset time window. The preset time window is a sliding time window with the current time as the endpoint. Calculating the real-time parking pressure value of each node based on the real-time operation data includes: calculating the difference between the entry count and the exit count corresponding to each node within the preset time window, as the net inflow count of each node within the preset time window; calculating the ratio of each net inflow count to each static capacity to obtain the net inflow rate of each node within the preset time window; and calculating the average of each net inflow rate and each real-time occupancy rate to obtain the real-time parking pressure value corresponding to each node.

[0009] Optionally, the step of performing pressure transmission simulation on each high-pressure node in the urban parking topology map based on each of the real-time weights to obtain the simulated parking pressure value of each node includes: in each transmission simulation, for each high-pressure node, performing the following steps: obtaining each potential node directly adjacent to the high-pressure node; using the real-time weights corresponding to the edges connecting the high-pressure node and each directly adjacent potential node as the distance attenuation coefficients corresponding to each directly adjacent potential node; calculating the product of each distance attenuation coefficient and the real-time parking pressure value of the high-pressure node to obtain the single transmission pressure value transmitted to the corresponding directly adjacent potential node; accumulating the single transmission pressure values ​​received by each potential node to the pressure buffer value of the corresponding potential node; when the transmission simulation is completed for the preset number of times, for each potential node, using the corresponding pressure buffer value as the total transmission pressure value of the corresponding potential node; and obtaining the simulated parking pressure value of each node based on the total transmission pressure value.

[0010] Optionally, obtaining the simulated parking pressure value of each node based on the total conduction pressure value includes: for each potential node, adding the corresponding real-time parking pressure value to the corresponding total conduction pressure value to obtain the simulated parking pressure value of the corresponding potential node; and for each high-voltage node, using the corresponding real-time parking pressure value as the simulated parking pressure value of the corresponding high-voltage node.

[0011] Optionally, the step of calculating the price adjustment coefficient corresponding to each target node based on the simulated parking pressure value corresponding to each target node includes: obtaining the maximum pressure value and the minimum pressure value among the simulated parking pressure values ​​in the potential congestion cluster; calculating the difference between the maximum pressure value and the minimum pressure value as a first difference; and performing the following steps for each target node to obtain the corresponding price adjustment coefficient: calculating the difference between the simulated parking pressure value and the minimum pressure value as a second difference; calculating the ratio of the second difference to the first difference to obtain the initial price adjustment coefficient of the target node; if the initial price adjustment coefficient is less than a preset lower limit, then setting the price adjustment coefficient to the preset lower limit; if the initial price adjustment coefficient is greater than a preset upper limit, then setting the price adjustment coefficient to the preset upper limit; if the initial price adjustment coefficient is greater than or equal to the preset lower limit and less than or equal to the preset upper limit, then using the initial price adjustment coefficient as the price adjustment coefficient of the target node.

[0012] Optionally, generating a diversion instruction based on the real-time parking rates includes: using the high-voltage node in the potential congestion cluster as the starting point of the diversion path; obtaining the real-time navigation travel time between the starting point and each potential node in the potential congestion cluster; calculating the ratio of each real-time navigation travel time to a preset benchmark travel time as the time factor corresponding to each potential node; calculating the product of the time factor and a first preset weight as the time cost value corresponding to each potential node; calculating the ratio of the real-time parking rate of each potential node to a preset benchmark rate as the rate factor corresponding to each potential node; calculating the product of the rate factor and a second preset weight as the rate cost value corresponding to each potential node, wherein the sum of the first preset weight and the second preset weight is 1; calculating the sum of the time cost value and the rate cost value to obtain the comprehensive cost value corresponding to each potential node; sorting the comprehensive cost values ​​from high to low, and selecting a preset number of potential nodes corresponding to the comprehensive cost values ​​from low to high, to generate the diversion instruction from the starting point to each selected potential node.

[0013] Secondly, embodiments of this application provide a city-level parking resource optimization system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, which includes computer instructions, and the one or more processors call the computer instructions to cause the city-level parking resource optimization system to perform the method described in the first aspect and any possible implementation thereof.

[0014] Thirdly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a city-level parking resource optimization system, cause the city-level parking resource optimization system to perform the method described in the first aspect and any possible implementation thereof.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a city-level parking resource optimization system, cause the city-level parking resource optimization system to perform the method described in the first aspect and any possible implementation thereof.

[0016] In summary, one or more technical solutions provided in this application have at least the following technical effects or advantages: 1. By constructing an urban parking topology map and integrating real-time operational and traffic data, high-pressure nodes and potential nodes are dynamically identified. Then, through pressure transmission simulation, potential congestion clusters that may lead to chain congestion are accurately predicted. Differentiated real-time parking rates are calculated for nodes within the clusters, and diversion instructions are generated. This achieves a leap from passive response to proactive prediction in urban-level parking management. It can systematically and proactively alleviate regional parking pressure, effectively avoid regional traffic paralysis caused by the oversaturation of a single parking lot, and significantly improve the accuracy of urban-level parking resource allocation.

[0017] 2. By using the reciprocal of the simulated travel time as the real-time weight of the edges, and combining the net inflow rate and real-time occupancy rate within the sliding time window to calculate the parking pressure value, and by integrating the real-time navigation travel time with the adjusted rate to generate a multi-factor cost-optimized diversion path, the characterization of traffic conditions and parking demand becomes more refined and dynamic. This improves the realism and accuracy of pressure transmission simulation and parking recommendations, ensuring that the generated diversion instructions not only consider spatial distance but also integrate real-time traffic conditions and users' economic costs, thereby significantly improving drivers' willingness to comply and the actual effectiveness of diversion measures.

[0018] 3. By simulating pressure transmission through multiple iterations based on real-time weights, the pressure from high-pressure nodes is diffused to surrounding potential nodes in a distance-attenuated manner. A method based on dynamic range normalization of simulated pressure values ​​is used to calculate the price adjustment coefficient, while upper and lower limits are set for constraint. This allows for a more scientific quantification of the congestion impact on surrounding nodes, enabling the development of smooth and reasonable price adjustment strategies. This avoids drastic price fluctuations, making rate adjustments more scientifically based and acceptable. It effectively guides traffic flow dispersion and prevents user complaints caused by price discrimination or unreasonable pricing, achieving a balance between economic and social benefits. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the city-level parking resource optimization method provided in an embodiment of this application; Figure 2 This is a schematic diagram of the process for determining the simulated parking pressure value provided in the embodiments of this application; Figure 3 This is a schematic diagram of a city-level parking resource optimization system provided in an embodiment of this application.

[0020] Explanation of reference numerals in the attached drawings: 701, Central Processing Unit; 702, Read-Only Memory; 703, Random Access Memory; 704, Bus; 705, Input / Output Interface; 706, Input Section; 707, Output Section; 708, Storage Section; 709, Communication Section; 710, Driver; 711, Removable Media. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] In the description of the embodiments of this application, words such as "illustrative," "for example," or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "illustrative," "for example," or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Rather, the use of words such as "illustrative," "for example," or "for example" is intended to present the relevant concepts in a specific manner.

[0023] In the description of the embodiments of this application, the terms "first" and "second" are used only to distinguish similar objects and do not represent a specific order of objects. It is understood that "first" and "second" can be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0024] In the embodiments of this application, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0025] In the implementation of this application, the collection and processing of relevant data should strictly comply with the requirements of relevant national laws and regulations, obtain the informed consent or separate consent of the personal information subject, and carry out subsequent data use and processing within the scope of laws and regulations and the authorization of the personal information subject.

[0026] Unless otherwise defined, all technical and scientific terms used in the embodiments of this application have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the embodiments of this application is for the purpose of describing the embodiments of this application only and is not intended to limit this application.

[0027] In related technologies, spatiotemporal prediction models are used to make relatively accurate predictions of parking space changes under normal conditions. Although existing GCN-based models can capture geographical proximity relationships, they cannot dynamically reflect the impact of real-time traffic conditions, which reduces the accuracy of city-level parking resource allocation. To address the above problems, this application provides a city-level parking resource optimization method and system, which can effectively improve the accuracy of city-level parking resource allocation.

[0028] Figure 1 This is a flowchart illustrating the city-level parking resource optimization method provided in this application embodiment.

[0029] This invention discloses a city-level parking resource optimization method, such as... Figure 1 As shown, the steps include the following.

[0030] S101. Obtain real-time operational data and geographical location of all parking lots within the target city area.

[0031] Specifically, by deploying sensor networks and data interfaces in all networked parking lots within the target urban area (such as an administrative district, a business district, or the entire city), the system continuously or periodically pulls and receives data. For real-time operational data, the system obtains the static attributes (such as the total number of parking spaces) and dynamic change attributes (such as the number of currently occupied parking spaces, the number of vehicles entering and exiting within a specific time period) from the central management system of each parking lot. This data is usually updated at a high frequency (e.g., once every minute or every five minutes). For geographic location data, the system obtains the precise latitude and longitude coordinates of the entrance or center point of each parking lot. This data usually remains unchanged after being entered once, unless the physical boundaries of the parking lot change. All these data items are cleaned, formatted, and correlated to generate a complete data record for each parking lot containing its unique identifier, real-time operational status, and precise geographic coordinates, laying a solid data foundation for subsequent construction of urban parking topology maps and analysis calculations.

[0032] The target city area refers to a geographical scope that this method intends to implement and optimize. This could be a heavily congested commercial center, a large residential area, or a specific area naturally formed by administrative divisions or transportation networks. Parking lots refer to all locations within the target city area that are included in the management system and available for vehicle parking, including but not limited to public parking lots, roadside parking spaces, parking lots attached to shopping malls and office buildings, and residential parking lots. Real-time operational data refers to a dynamic set of information reflecting the current operational status and short-term trends of parking lots. Its core characteristic is that it is real-time or near real-time, used to quantify the busyness and supply-demand situation of parking lots. Geographical location refers to information used to uniquely identify the location of a parking lot in geographic space. It is usually represented by Global Positioning System (GPS) coordinates or National Geodetic Coordinate System (GMO) coordinates (such as CGCS2000) coordinates and is crucial for spatial analysis and the construction of topological relationships.

[0033] S102. Construct a city parking topology with each parking lot as a node and the connecting roads between each node as edges.

[0034] Specifically, using the precise geographic location information of all parking lots obtained in step S101, they are marked as independent point features in a digital map or geographic information system (GIS), with each point representing a node. Based on urban road network data, drivable road paths connecting any two parking lot nodes are found and determined. These paths are abstracted as edges connecting nodes, not simply using straight-line distances as edges, but based on real, drivable roads (such as urban main roads, auxiliary roads, one-way streets, etc.) for path planning. Finally, a network graph consisting of a set of nodes and a set of edges is formed. This graph fully reflects the spatial connectivity between all parking lots within the target urban area, providing a core data structure for subsequent complex graph-based calculations (such as pressure transmission and shortest path analysis).

[0035] In graph theory, a node represents a specific parking lot. Each node carries the parking lot's attribute information, such as its ID, name, and real-time operational data, serving as the basic unit for computation and state representation within the network. An edge, another fundamental concept, represents the actual road connecting two parking lot nodes. An edge links two nodes, indicating that vehicles can travel directly or indirectly between the two parking lots via this road. A connecting road, in a real urban road network, refers to a complete driving route that connects the entrances or exits of two parking lots. It may consist of multiple road segments but is abstracted as a single edge in the graph. The urban parking topology graph is the final network model describing the connectivity between parking lots. It focuses on the connections between nodes rather than precise geometry (such as curvature), capturing the spatial distribution and interrelationships of parking resources.

[0036] S103. Based on real-time traffic data of the target city area, obtain the real-time weights of each side.

[0037] Specifically, the system accesses real-time traffic data streams from the target city area. These data may come from sensors in the transportation department, floating car GPS data, or real-time traffic interfaces of map service providers. No specific restrictions are imposed here. Each edge (i.e. connecting road) in the topology map is matched with the corresponding road segment in the real-time road network. Then, based on the real-time traffic data, the current traffic capacity of the road segment is calculated or directly obtained as the real-time weight. This weight value is updated periodically to ensure that it reflects the latest traffic conditions.

[0038] Real-time traffic data refers to a dataset that dynamically reflects the current operational status of the urban road network, used to represent the smoothness and congestion of traffic flow. Its sources include, but are not limited to, inductive loop detectors, cameras, taxi GPS trajectories, and crowdsourced data from navigation software. Real-time weights are dynamically changing values ​​assigned to each edge, used in graph computation to characterize the cost, burden, or capacity required for that edge. Their core characteristic is real-time updates, reflecting the dynamic nature of traffic conditions.

[0039] Through the above embodiments, this method upgrades static road connections into dynamic connections that can sensitively respond to the pulse of urban traffic, assigns real-time weights to each edge, and realizes the evolution of the urban parking topology map from a simple connectivity model to a highly realistic dynamic network model containing toll cost information. This improves the accuracy and practicality of stress simulation and subsequent resource optimization strategies, and ensures that the system's suggestions are highly consistent with the driver's real experience and the objective conditions of the road.

[0040] Based on the above embodiments, as an optional embodiment, for Figure 1 The step S103 shown can be implemented through steps S201-S203, which will be explained in detail below.

[0041] S201. Obtain the actual traffic distance and preset traffic speed for each side.

[0042] Specifically, for actual traffic distance, based on the constructed urban parking topology map, the system calls the path planning interface of a Geographic Information System (GIS) or high-precision digital map to obtain the optimal path length represented by each edge along the actual passable road network between two parking nodes. This distance is relatively static and only needs to be updated when the road network undergoes permanent changes. For preset traffic speed, the system deeply integrates real-time traffic data. It does not simply adopt road speed limits, but comprehensively analyzes the current time, historical average vehicle speed, real-time floating vehicle (such as taxis and logistics vehicles) GPS trajectory speed, and traffic event information (such as accidents and construction). Through data fusion and machine learning models, it dynamically predicts the representative driving speed that vehicles can actually reach on that road at the current moment and in the short future. Finally, the system accurately binds an actual traffic distance value and a dynamically updated preset traffic speed value to each edge in the topology map.

[0043] The actual traffic distance refers to the total length of the shortest or optimal path that a vehicle can travel from the road near the entrance of the starting parking lot to the road near the entrance of the ending parking lot in a real urban road network, according to traffic rules. It is used to represent the spatial displacement cost between two points. The preset traffic speed is a dynamic evaluation value used to characterize the average speed that a vehicle is expected to reach when traveling on the road corresponding to a certain edge under current traffic conditions. It is a real-time quantification of road capacity and a key bridge for converting spatial distance into time cost.

[0044] S202. Calculate the ratio of each actual traffic distance to the preset traffic speed to obtain the simulated travel time for each side.

[0045] Specifically, for each edge in the city parking topology map, the prepared actual traffic distance value and preset traffic speed value are read from its associated data structure. The actual traffic distance is used as the dividend and the preset traffic speed is used as the divisor to perform division calculation. The result is the average time that a vehicle is expected to take to travel through the road path represented by that edge under the current traffic conditions, which is the simulated travel time. Each edge is assigned a calculated simulated travel time value, which will be used as an intermediate variable for the real-time weight of subsequent edge generation.

[0046] The simulated travel time is the output of this calculation step. It is a numerical value representing the time cost and is used to estimate the length of time required to travel along an edge. Its unit is usually hours or minutes.

[0047] S203. Calculate the reciprocal of each simulated passage time to obtain the real-time weight of each edge.

[0048] Specifically, after obtaining the simulated travel time for each edge, it is used as input to perform a mathematical reciprocal operation (i.e., 1 divided by the simulated travel time). The shorter the simulated travel time of an edge (indicating smooth traffic and high efficiency), the larger its reciprocal—that is, its real-time weight—is; conversely, the longer the simulated travel time (indicating congested traffic and low efficiency), the smaller its real-time weight is. Through this transformation, the travel time in the physical world is mapped to the connection strength in the graph theory network. Edges with shorter travel times are considered strong connections with high weights, while edges with longer travel times are considered weak connections with low weights. Ultimately, each edge in the topology graph is updated with a weight value representing its current travel efficiency.

[0049] S104. Based on the real-time operation data, calculate the real-time parking pressure value of each node. The real-time operation data shall include at least the static capacity and real-time occupancy rate of the corresponding parking lot.

[0050] Specifically, real-time operational data should at least characterize the static capacity of the parking lot's maximum capacity to accommodate vehicles, and the real-time occupancy rate reflecting the current availability of parking spaces. In addition to these two basic data points, for a more accurate assessment, supplementary data such as real-time vehicle entry rate, exit rate, average parking time, surrounding road congestion index, and data related to special events (e.g., large-scale events) can be collected. This multi-dimensional, heterogeneous real-time operational data is then input into a pre-defined mathematical model or algorithm. This model assigns appropriate weights to different data indicators and performs comprehensive calculations to quantify the ease of finding a parking space at a given node in the current and foreseeable future period into a specific numerical value. This value is the real-time parking pressure value for that node. Ultimately, each node will obtain a dynamically updated pressure value for subsequent analysis or decision-making.

[0051] The real-time parking pressure value, an output of this step, is a comprehensive quantitative indicator used to characterize the overall intensity of parking demand load a parking lot is experiencing at a given moment. A higher value indicates that the parking lot is under strain and is closer to full capacity or oversaturation. Static capacity refers to the maximum number of vehicles a parking lot can accommodate in its design. It is a relatively fixed attribute and is used as the denominator in calculations to standardize absolute values ​​(such as the number of occupied vehicles or net inflow vehicles) into relative ratios, making parking lots of different sizes comparable. Real-time occupancy rate refers to the percentage of occupied parking spaces in the parking lot relative to the total static capacity at the current moment. It is an instantaneous state indicator used to directly reflect the current space utilization saturation level of the parking lot.

[0052] Based on the above embodiments, as an optional embodiment, the real-time operation data also includes the entry count and exit count of the corresponding node within a preset time window. The preset time window is a sliding time window with the current time as the endpoint. Figure 1 The step S104 shown can be implemented through steps S301-S303, which will be explained in detail below.

[0053] S301. Calculate the difference between the inflow count and the outflow count of each node within a preset time window, and use it as the net inflow count of each node within the preset time window.

[0054] Specifically, each node maintains a fixed-duration window (e.g., the past 15 minutes) that extends forward from the current time. Within this window, vehicle passage events from the parking lot entrance and exit sensors are continuously accumulated. The total entrance count (i.e., the total number of vehicles entering) and the total exit count (i.e., the total number of vehicles leaving) within this window are recorded. The exit count is subtracted from the entrance count, and the result is the net inflow count for that node within the specified time window. This value is a signed integer, and its sign and magnitude directly reflect whether the parking lot has experienced a net inflow of vehicles (positive value, indicating increased demand and pressure) or a net outflow (negative value, indicating easing demand and reduced pressure) in the recent period, as well as the severity of this change.

[0055] The preset time window refers to a fixed-length time interval whose endpoint is always aligned with the current time. It is a sliding window that continuously discards older data (such as instantaneous counts from 15 minutes ago) and incorporates the latest data over time, thus always reflecting the state of the most recent period. The entrance count refers to the total number of vehicles passing through the parking lot's entrance gates or detection equipment within the preset time window; it is an absolute value representing the inflow of parking demand. The exit count refers to the total number of vehicles passing through the parking lot's exit gates or detection equipment within the preset time window; it is an absolute value representing the outflow of parking demand. The net inflow count is a signed integer value. A positive value indicates that more vehicles entered than left during the window period, resulting in a net increase in the number of vehicles in the parking lot. A negative value indicates that more vehicles left than entered, resulting in a net decrease in the number of vehicles in the parking lot. A zero value indicates a balance between inflow and outflow.

[0056] S302. Calculate the ratio of each net inflow count to each static capacity to obtain the net inflow rate of each node within the preset time window.

[0057] Specifically, the net inflow count calculated in step S301 for each node and the inherent static capacity value of that node are read from the internal storage. The net inflow count is used as the dividend and the static capacity is used as the divisor. The simple change in the number of vehicles is transformed into the rate of load change relative to the maximum carrying capacity of the parking lot. A standardized net inflow rate is output for each node. This value, in the form of a ratio or percentage, intuitively reflects the relationship between the degree of increase in the parking lot load and its own size within a preset time window.

[0058] The net inflow rate is a dimensionless ratio that represents the proportion of the net increase in the number of vehicles in a parking lot to its total capacity within a preset time window. It is used to standardize the rate of change of parking lot load.

[0059] S303. Calculate the average of each net inflow rate and each real-time occupancy rate to obtain the real-time parking pressure value corresponding to each node.

[0060] Specifically, the arithmetic mean of the net inflow rate and real-time occupancy rate is calculated by adding them together and then dividing by two. This calculation process gives equal weight to the current state and future trends, considering both equally important in determining the overall pressure level of a node. A single real-time parking pressure value is output for each node, which comprehensively represents the node's current congestion level and the risk of further deterioration in the near future, thus providing a more comprehensive and forward-looking load view than any single indicator.

[0061] S105. Nodes with real-time parking pressure values ​​greater than or equal to the first preset pressure value are identified as high-pressure nodes, and nodes with real-time parking pressure values ​​less than the first preset pressure value are identified as potential nodes.

[0062] Specifically, based on the calculated real-time parking pressure value, all nodes in the city parking topology map are clearly divided into two sets with different characteristics and functions. The real-time parking pressure value of each node is compared one by one with a predefined and set threshold—the first preset pressure value. If the pressure value of a node is greater than or equal to the threshold, the system marks it and classifies it as a high-pressure node. Conversely, if its pressure value is strictly less than the threshold, it is marked and classified as a potential node. After traversing and judging all nodes, the system obtains a list of high-pressure nodes and a list of potential nodes. This lays the foundation for completely different processing logics (high-pressure nodes act as pressure sources for pressure transmission, while potential nodes act as potential pressure bearers and destinations for traffic diversion).

[0063] The first preset pressure value is a fixed threshold set by the system administrator or algorithm based on historical data, urban parking management goals, and experience. It serves as a benchmark to determine whether the pressure at a node has reached a high-risk level requiring special attention. High-pressure nodes are those whose real-time parking pressure values ​​have reached or exceeded the danger threshold. These nodes indicate that they are currently in or about to enter an oversaturated state, and are direct risk points and sources of traffic congestion and parking difficulties. They require immediate attention and serve as the starting point for pressure transmission. Potential nodes are those whose real-time parking pressure values ​​have not yet reached the danger threshold. These nodes indicate that the current situation is relatively relaxed, with a certain amount of available capacity. However, they are not insignificant; rather, they serve as reserve resources and safety buffers to absorb vehicles diverted from high-pressure nodes and alleviate regional congestion.

[0064] S106. Based on the real-time weights, pressure transmission simulation is performed on each high-pressure node in the urban parking topology map to obtain the simulated parking pressure value of each node.

[0065] Specifically, the simulation demonstrates how pressure spreads to neighboring nodes when a high-pressure node can no longer accept new parking demands. The amount and direction of pressure transmission are not uniformly distributed but are strictly allocated according to real-time weights—pressure flows preferentially and in greater quantities to nodes connected via paths with lower weights (i.e., smoother and more convenient passage). This transmission process is iterative. When the pressure value of a neighboring node that has received pressure also accumulates to a certain threshold, it will continue to transmit pressure to its surrounding nodes. This cycle continues until the pressure distribution in the entire topology reaches a new equilibrium or reaches a preset number of iterations. After this series of complex chain reaction simulations, the initial pressure values ​​of the initial high-pressure node and all its affected nodes in the diagram are updated, thus calculating and outputting a new quantitative indicator, namely the simulated parking pressure value, for each node.

[0066] Among them, pressure transmission simulation is a graph theory-based algorithmic metaphor that abstracts the spatiotemporal correlation of traffic congestion into a computable and diffuse pressure in a network graph. The simulated parking pressure value is a node-level attribute that characterizes the comprehensive and predictive pressure level faced by the node after considering its own state and the cascading effects from the network.

[0067] Figure 2 This is a schematic diagram of the process for determining the simulated parking pressure value provided in the embodiments of this application.

[0068] Based on the above embodiments, as an optional embodiment, for Figure 1 The step S106 shown can be achieved through... Figure 2 The steps S401-S406 are implemented, and will be explained in detail below.

[0069] In each conduction simulation, steps S401-S404 are performed for each high-voltage node: S401. Obtain all potential nodes directly adjacent to the high-voltage node.

[0070] Specifically, the system iterates through the categorized list of high-pressure nodes. For each high-pressure node in the list, the system queries the pre-built urban parking topology data structure. Based on the connection relationships (edges) in the graph, it retrieves all other nodes that are directly connected to the high-pressure node. The system then performs an intersection operation on these retrieved neighbor nodes and the categorized list of potential nodes, filtering out nodes that simultaneously meet the conditions of being both a neighbor node and a potential node. These filtered nodes are the potential nodes directly adjacent to the high-pressure node, and they will serve as the direct target objects receiving pressure from the high-pressure node in this round of pressure transmission simulation.

[0071] S402. Use the real-time weights of the edges connecting the high-voltage node and each directly adjacent potential node as the distance attenuation coefficients for each directly adjacent potential node.

[0072] Specifically, after obtaining all directly adjacent potential nodes for a given high-pressure node, parameter mapping is performed for each such potential node-high-pressure node pair. From the city parking topology map, the edge connecting this pair of nodes is precisely located, and the real-time weight of that edge is read. This real-time weight value is directly assigned to the directly adjacent potential node acting as the pressure receiver. However, the role and semantics of this value change at this point; it is defined as the distance attenuation coefficient for the potential node when receiving pressure from this specific high-pressure node. This coefficient will be directly used to calculate the pressure value that is about to be transmitted.

[0073] Among them, the distance attenuation coefficient corresponding to each directly adjacent potential node is a coefficient specific to the transmission path from a specific high-pressure node to a specific potential node. It is used in the subsequent pressure calculation to quantitatively attenuate the original pressure value transmitted from the high-pressure node, reflecting the weakening or strengthening effect of road traffic conditions on pressure propagation.

[0074] S403. Calculate the product of each distance attenuation coefficient and the real-time parking pressure value of the high-pressure node to obtain the single transmission pressure value transmitted to the corresponding directly adjacent potential node.

[0075] Specifically, when processing a specific high-voltage node, the system iterates through each directly adjacent potential node and its assigned distance attenuation coefficient. For each potential node-attenuation coefficient pairing, the system performs a multiplication operation: the distance attenuation coefficient is used as the multiplier and multiplied by the real-time parking pressure value of the high-voltage node, which is the multiplicand. The original pressure intensity of the high-voltage node is scaled according to the real-time traffic efficiency of the connecting road (i.e., the attenuation coefficient). The result is the single transmission pressure value from the high-voltage node to the specific potential node in this transmission simulation. This value represents the intensity of the direct and immediate impact of the high-voltage node on its neighbors under the current network conditions.

[0076] The single transmission pressure value represents the amount of pressure transmitted from the source node to the target node along a specific edge in one simulation iteration.

[0077] S404. Accumulate the single transmission pressure value received by each potential node to the pressure buffer value of the corresponding potential node.

[0078] Specifically, in each round of conduction simulation, when a potential node receives multiple single conduction pressure values ​​from multiple different high-voltage nodes, a temporary data storage, called a pressure cache value, is maintained for each potential node. All potential nodes are traversed, and for each potential node, all single conduction pressure values ​​received from all sources (i.e., all directly connected high-voltage nodes) in this round of simulation are found. Then, an addition operation is performed to sum all these values. This summation result does not overwrite the original cache value, but is added to it, that is, added to the pressure cache value that already exists after the previous round of simulation for that potential node, thereby updating the cache value. This continuously increasing pressure cache value will eventually be used to characterize the total external conduction pressure that the potential node has experienced throughout the entire multi-round simulation process.

[0079] The pressure cache value is a temporary accumulator variable allocated by the system for each potential node. It is usually initialized to 0 before the simulation begins, and then in each round of simulation, it is responsible for accumulating and recording the sum of all transmitted pressures received by the node from the network.

[0080] S405. When the preset number of conduction simulations is completed, for each potential node, the corresponding pressure buffer value is used as the total conduction pressure value of the corresponding potential node.

[0081] Specifically, when the simulation starts, a preset number of iterations is set. This number determines the range and depth of pressure propagation in the network. The aforementioned pressure allocation and accumulation operations continue until the number of simulation iterations reaches this preset value. At this point, the simulation stops, and no new pressure values ​​are generated or transmitted. All nodes marked as potential nodes are traversed, and the final value stored in each node's pressure cache value temporary accumulator is read. This value represents the sum of all transmitted pressure received by the potential node from all high-pressure nodes in the network and other nodes that may indirectly transmit pressure (in multi-round simulations, pressure can jump through multiple nodes) throughout all preset simulation iterations. This final pressure cache value is formally and once assigned to the node and identified as the node's total transmitted pressure value. This value will then be used as a fixed attribute in subsequent calculations.

[0082] The preset number of simulations is a predefined integer that controls the scale of the pressure transmission simulation. The more simulations, the farther the pressure propagates and the more nodes are affected, but the computational load also increases. The total transmitted pressure value quantitatively describes the total amount of external transmitted pressure from the entire parking lot network that a potential node experiences.

[0083] For example, assume the system is set to run 3 rounds. Before the simulation begins, the pressure buffer value of all potential nodes is initialized to 0. After the first round of simulation, the pressure buffer value of potential node P accumulates to 25%. After the second round of simulation, the pressure buffer value of node P accumulates by another 15%, becoming 40%. After the third round of simulation, the pressure buffer value of node P accumulates by another 10%, becoming 50%. At this point, the 3 rounds of simulation are complete. For potential node P, its pressure buffer value of 50% is taken as its total transmitted pressure value. This means that the system determines that node P has been affected by external congestion equivalent to 50% of its own pressure baseline.

[0084] S406. Based on the total transmission pressure value, the simulated parking pressure value of each node is obtained.

[0085] Specifically, two core data points are integrated: first, the inherent pressure of the node itself, i.e., the real-time parking pressure value, representing its current parking load; and second, the total transmitted pressure value just calculated, representing the potential future parking demand from the surrounding environment. These two pressures are superimposed or weighted to generate a new, comprehensive quantitative indicator for each node, namely the simulated parking pressure value (no specific restrictions are placed here). This final value not only reflects the current saturation level of the parking area but also proactively includes the additional parking demand that may surge due to the shortage of parking spaces in the surrounding area, thus more accurately predicting the degree of parking difficulty in the near future.

[0086] Based on the above embodiments, as an optional embodiment, for Figure 2 The illustrated step S406 can be achieved through... Figure 2 Steps S4061-S4062 are implemented, and are explained in detail below. S4061. For each potential node, add the corresponding real-time parking pressure value to the corresponding total transmission pressure value to obtain the simulated parking pressure value of the corresponding potential node.

[0087] Specifically, all nodes classified as potential nodes are traversed. For each such node, two key values ​​are read from the data storage: one is the real-time parking pressure value representing its internal load status, and the other is the total transmission pressure value representing all external influences received from the entire parking network. These two values ​​are added together to combine the static or current tension of a parking lot with the dynamic or future input risks that may be caused by congestion in its surrounding environment. The result of this addition is the final simulated parking pressure value of the potential node. This new value provides a more comprehensive and predictive perspective than the original real-time pressure value to assess the true risk level of the node in the regional parking network.

[0088] Through the above embodiments, accurate reassessment and early warning of potential node risks are achieved. By adding internal pressure to externally transmitted pressure, the system can reveal high-risk points that appear relaxed on the surface but are actually fraught with hidden dangers, improving the accuracy and timeliness of parking resource optimization.

[0089] S4062. For each high-voltage node, the corresponding real-time shutdown pressure value is used as the simulated shutdown pressure value of the corresponding high-voltage node.

[0090] Specifically, for nodes that have been identified as high-pressure nodes, external transmission pressure values ​​are no longer introduced for superposition. Instead, the inherent real-time parking pressure value of each high-pressure node is directly assigned to the simulated parking pressure value attribute of that node. The final simulated value of a high-pressure node is entirely determined by its own internal state. The system considers it as the source of pressure, and its own congestion level is already high enough. There is no need to further emphasize its severity by superimposing relatively minor transmission pressures that may come from other nodes. This operation ensures that the pressure identification of high-pressure nodes is clear and direct, and consistent with its initial criteria.

[0091] Through the above embodiments, the source-first logic simplification principle is followed when processing high-pressure nodes. By directly using real-time pressure values ​​as simulated pressure values, the status identification of high-pressure nodes is ensured to be stable and easy to trace, avoiding unnecessary fluctuations or dilution of their core status due to complex network feedback effects.

[0092] S107. Potential nodes whose simulated parking pressure values ​​are greater than the second preset pressure value and less than the first preset pressure value are identified together with a corresponding high-pressure node as a potential congestion cluster. The second preset pressure value is less than the first preset pressure value, and the actual traffic distance between the potential node and the corresponding high-pressure node is less than a preset distance threshold.

[0093] Specifically, all potential nodes are traversed, and these nodes are finely screened using two preset pressure thresholds: a first preset pressure value (used to define high-pressure nodes) and a second preset pressure value (a lower threshold). Potential nodes whose simulated parking pressure values ​​fall between these two thresholds are selected. These nodes are considered high-risk potential nodes that have been significantly affected and have a higher risk. For each selected high-risk potential node, the corresponding high-pressure node that exerts a major influence on it is found. This is achieved through the principle of spatial proximity, i.e., checking whether the actual traffic distance between the high-risk potential node and each high-pressure node is less than a preset distance threshold. If a high-risk potential node and a high-pressure node meet this distance condition, they are bound together. Ultimately, at least one such high-risk potential node and its bound high-pressure node together constitute a potential congestion cluster. This cluster represents a set of parking lots that are spatially proximate, interconnected in terms of pressure state, and have a high risk of collaborative congestion.

[0094] The second preset pressure value is a lower threshold set by the system to identify high-risk potential nodes, defining which potential nodes require focused attention. The preset distance threshold is a spatial limitation set by the system to ensure that nodes within a cluster are physically capable of collaborative management. A potential congestion cluster is a logical grouping containing a high-pressure node and one or more high-risk potential nodes directly affected by it, representing a potential area of ​​collaborative congestion.

[0095] Through the above embodiments, a strategic upgrade from discrete node management to regional collaborative management has been achieved. By defining potential congestion clusters through the two dimensions of joint pressure level and spatial distance, it is possible to accurately identify parking lot groups that are physically adjacent, mutually influential in state, and may jointly cause regional traffic paralysis. This improves the efficiency and effectiveness of intervention measures and effectively prevents the spread of local congestion to a wider area.

[0096] For each potentially congested cluster, execute steps S108-S110 to optimize resources.

[0097] S108. Based on the simulated parking pressure value corresponding to each target node, calculate the price adjustment coefficient corresponding to each target node. The target node is any node in the potential congestion cluster.

[0098] Specifically, the process iterates through all identified potential congestion clusters, focusing on any node belonging to one of these clusters. These nodes become the target nodes for this price adjustment calculation. The simulated parking pressure value of each target node is used as the core input variable and substituted into a pre-defined function mapping model. This model aims to transform the dimensionless pressure value into a price adjustment coefficient with practical economic significance. This function model is usually designed to be piecewise or non-linear. When the simulated parking pressure value of a target node is below a certain mitigation threshold, its output price adjustment coefficient may be 1.0, meaning that the original price is maintained. Once the pressure value exceeds the threshold and continues to rise, the price adjustment coefficient will increase accordingly, and the growth slope may become increasingly steep, for example, gradually increasing from 1.1 to 1.5 or higher. No specific restrictions are imposed here. Finally, a precise price adjustment coefficient directly related to its future congestion risk is generated and output for each selected target node.

[0099] The target node refers to any node belonging to the currently being processed potential congestion cluster. It can be either a core high-pressure node within the cluster or a high-risk potential node that has been drawn into the cluster. The price adjustment coefficient is a multiplier applied to the base rate. A coefficient greater than 1 indicates an increase in price (intended to suppress demand or guide outflow), while a coefficient less than 1 indicates a decrease in price (intended to attract demand or guide inflow).

[0100] Through the above embodiments, a precise transition from state perception to economic leverage regulation is achieved. By calculating the price adjustment coefficient based on simulated parking pressure values, abstract congestion risks and pressure levels can be translated into specific and actionable price signals. This enables intelligent redistribution of parking demand and is one of the most core and effective economic means to mitigate potential congestion cluster risks and achieve resource optimization.

[0101] Based on the above embodiments, as an optional embodiment, for Figure 1 The step S108 shown can be implemented through steps S501-S507, which will be explained in detail below.

[0102] S501. Obtain the maximum and minimum pressure values ​​among the simulated parking pressure values ​​in the potential congestion cluster.

[0103] Specifically, when processing a particular potential congestion cluster, all target nodes contained in the cluster (i.e., each parking lot node belonging to the cluster) are traversed, and their simulated parking pressure values ​​are read. These values ​​are temporarily stored as a data set. A search operation is performed on this pressure value set. By comparing the values, the maximum and minimum values ​​are found and recorded. These two extreme values ​​will serve as key baseline parameters for subsequent normalization processing and price adjustment coefficient calculation for each node in the cluster.

[0104] The maximum pressure value, the largest value in this dataset, indicates the upper limit of shutdown pressure within the cluster, typically corresponding to the core high-pressure node within the cluster. The minimum pressure value, the smallest value in this dataset, indicates the lower limit of shutdown pressure within the cluster, typically corresponding to a high-risk potential node within the cluster that is relatively less affected. Through the above embodiments, this method lays a scientific quantitative foundation for subsequent differentiated pricing.

[0105] S502. Calculate the difference between the maximum pressure value and the minimum pressure value, and use it as the first difference value.

[0106] Specifically, after successfully obtaining the maximum and minimum pressure values ​​of the currently processed potential congestion cluster, the minimum pressure value is subtracted from the maximum pressure value (as the minuend). The result, the numerical difference between the two, is defined by the system as the first difference. This first difference represents the entire range of pressure change from the least stressed node to the most stressed node within the cluster, reflecting the magnitude of the stress gradient in the parking pressure distribution within the cluster. This value will serve as the core denominator in subsequent normalization calculations, scaling the absolute pressure difference of each node to a uniform relative scale.

[0107] For each target node, steps S503-S507 are executed to obtain the corresponding price adjustment coefficient: S503. Calculate the difference between the simulated parking pressure value and the minimum pressure value, and use it as the second difference value.

[0108] Specifically, the process iterates through each target node within the cluster. For the currently being processed target node, its simulated parking pressure value is read. The minimum pressure value of the cluster is used as the subtrahend and subtracted from the simulated parking pressure value of that node (used as the minuend). The result is the numerical difference between the node's pressure and the lowest pressure in the cluster, which is defined by the system as the second difference value corresponding to that node. The second difference value is a non-negative value that intuitively represents how much more stressed the node is compared to the least stressed node in the cluster; it is an absolute measure of the additional pressure it bears. This value will serve as the main numerator in the subsequent calculation of its price adjustment coefficient.

[0109] S504. Calculate the ratio of the second difference to the first difference to obtain the initial price adjustment coefficient of the target node.

[0110] Specifically, the second difference is used as the dividend, and the first difference is used as the divisor. The result is the initial price adjustment coefficient for that target node. Mathematically, this coefficient represents the relative proportion of the node's pressure within the total pressure range of the cluster. If the coefficient is 0, it indicates that the node is at the lowest pressure point within the cluster; if it is 1, it indicates that the node is at the highest pressure point within the cluster. This value forms the direct basis for implementing differentiated price adjustments.

[0111] The initial adjustment factor is a theoretical value between 0 and 1, which quantitatively describes the exact position of node pressure in the continuous spectrum from the lightest to the heaviest within the cluster.

[0112] Through the above embodiments, a precise and fair mapping from pressure data to economic adjustment parameters is achieved through a simple ratio calculation, ensuring that the final pricing strategy can effectively reflect internal differences and has mathematical rigor and logical consistency.

[0113] S505. If the initial price adjustment coefficient is less than the preset lower limit, then the price adjustment coefficient is set to the preset lower limit.

[0114] Specifically, after calculating the initial price adjustment coefficient for a target node, it is immediately compared with a pre-configured threshold known as the preset lower limit. If the initial price adjustment coefficient is lower than this lower limit, the system will not use this excessively low calculated value. Instead, it will directly set the node's final price adjustment coefficient to this preset lower limit. This step ensures that for a node, no matter how low its calculated initial coefficient is, its final price adjustment coefficient will not fall below the minimum level allowed by the management strategy.

[0115] The preset lower limit is a fixed numerical threshold set in advance by the management system, which defines the bottom line for the price reduction. It is usually determined based on factors such as operating costs, market acceptance, or policy requirements.

[0116] S506. If the initial price adjustment coefficient is greater than the preset upper limit, then the price adjustment coefficient is set to the preset upper limit.

[0117] Specifically, if the initial price adjustment coefficient is greater than the upper limit, the system determines that the calculation result exceeds the allowed adjustment range, performs an overwrite operation, ignores the excessively high calculated value, and directly sets the final price adjustment coefficient of the node to the preset upper limit, ensuring that the final price adjustment coefficient of any node will not exceed the highest ceiling set by the management strategy.

[0118] The preset upper limit is a fixed numerical threshold set in advance by the management, which defines the maximum range of price increases allowed by law or policy. It is usually determined based on users' affordability, price regulatory requirements, or social impact assessment.

[0119] The above embodiments introduce crucial constraints and safeguards to the dynamic pricing mechanism. By enforcing the upper limit, the system effectively prevents exorbitant parking fees that may occur under extreme congestion conditions, thus avoiding public controversy and negative social impact.

[0120] S507. If the initial price adjustment coefficient is greater than or equal to the preset lower limit and less than or equal to the preset upper limit, then the initial price adjustment coefficient shall be used as the price adjustment coefficient of the target node.

[0121] Specifically, when the initial coefficient falls entirely within the closed interval defined by the lower and upper limits, the system determines that the initial coefficient is legal and valid. Without any truncation, it directly performs an assignment operation, using this unmodified initial adjustment coefficient as the final adjustment coefficient for that node.

[0122] The above embodiments ensure the flexibility and accuracy of the pricing strategy within the safety boundary. The system respects and adopts the majority of initial calculation results that fall within the reasonable range. This complements the steps for handling out-of-bounds situations and together constitutes a complete, robust, and flexible coefficient determination logic. This ensures that the entire resource optimization method will not lose its sensitivity due to excessive constraints, nor will it become uncontrollable due to a lack of constraints.

[0123] S109. Based on the price adjustment coefficient, obtain the real-time parking fee rate after price adjustment for each target node.

[0124] Specifically, after determining the final price adjustment coefficient for each target node within a potential congestion cluster, the benchmark parking rate applicable to each target node is obtained. This benchmark rate may be the normal price of the parking lot during non-congestion periods, or it may be a unified reference price set by the management department. The benchmark parking rate of the node is used as the multiplicand, and its price adjustment coefficient is used as the multiplier. The product of the two is the real-time parking rate after the price adjustment for that node. This newly calculated rate will be immediately updated in the parking lot's charging system and published to the public through external electronic price displays, mobile applications, and other means, thereby influencing drivers' parking choices through price signals.

[0125] The adjusted real-time parking rate is a specific, soon-to-be-implemented price, typically expressed as yuan per hour or yuan per 15 minutes.

[0126] S110. Generate traffic diversion instructions based on real-time parking rates.

[0127] Specifically, when a parking request (such as a user query from a navigation app) is triggered, the system immediately obtains the user's current geographical location and scans all parking nodes within a certain radius of that location. For each potential parking node considered, the system instantly retrieves its latest real-time parking rate. The key to generating the optimal instruction lies in a multi-dimensional comprehensive evaluation: not only comparing prices, but also simultaneously integrating the simulated parking pressure value of each node (predicting its future congestion risk) and the real-time travel cost from the user's current location to that node (such as the estimated driving time, this cost is determined by the city). The system, determined by real-time weights in the city's traffic topology map, calculates a referral score for each potential parking option using a comprehensive utility function. This score represents the optimal balance between price, convenience (time / distance), and reliability (low congestion risk). Options with excessively low scores (such as high prices, long distances, or exceeding predicted pressure limits) are eliminated. The system then combines the highest-scoring parking nodes with their optimal routes to form a clear and actionable set of recommendations. This comprehensive set of recommendations, including specific destinations, predicted costs, estimated travel time, and recommended routes, constitutes the final referral instruction. Its core objective is to proactively guide drivers away from parking areas that currently appear to have available spaces but are nearing saturation or have low cost-effectiveness, and to distribute parking demand to areas with the highest overall utility.

[0128] The traffic diversion command is not a single command, but a set of information, which usually includes a recommended starting area (around high-voltage nodes), a recommended destination parking lot (one or more low-voltage nodes), a real-time optimized route to the destination, and real-time rates and availability information for the destination parking lot. The core purpose is to guide traffic flow.

[0129] Through the above embodiments, a closed-loop management system from static price adjustment to dynamic route guidance is achieved. It not only tells drivers where the prices are cheap (price signal), but also directly tells them how to get there (route guidance), and provides decision support on whether it is worthwhile to go (comprehensive cost assessment). It connects scattered, independent parking lots into a collaborative network through information and routes, which can proactively, quickly, and efficiently disperse traffic pressure from congested points to idle resource points, thereby balancing parking demand in space and time, effectively alleviating regional traffic congestion, and maximizing the overall utilization efficiency of urban parking resources.

[0130] Based on the above embodiments, as an optional embodiment, for Figure 1 The step S110 shown can be implemented through steps S601-S608, which will be explained in detail below.

[0131] S601, use the high-voltage node in the potential congestion cluster as the starting point of the diversion path.

[0132] Specifically, when dealing with an identified potential congestion cluster, parking lots marked as high-voltage nodes are explicitly designated from the nodes contained in the cluster as the unified starting point for all diversion path calculations. Instead of starting from any point in the city, the navigation starting point is precisely anchored to these high-voltage nodes themselves or their adjacent entrance roads.

[0133] The diversion path refers to the vehicle route designed and recommended to alleviate traffic congestion from a congested point to an uncongested point. The starting point of the path refers to the spatial location at the beginning of a navigation path, which is specifically specified as the geographical coordinates of the high-voltage node in this embodiment.

[0134] S602. Obtain the real-time navigation passage time between the starting point of the path and each potential node in the potential congestion cluster.

[0135] Specifically, the process iterates through every potential node within the current potential congestion cluster, treating these nodes sequentially as the endpoints of the path. For each start-end pairing, it calls an external real-time navigation engine or map service interface, inputting the precise geographical locations of the start and end points, and requesting the driving time calculated based on the current real-time traffic conditions. This time is not based on the theoretical value of static distance and speed limit, but rather an estimated time derived by comprehensively considering dynamic factors such as real-time traffic congestion, traffic light waiting, and road construction events. This provides each potential node within the cluster with a real-time navigation travel time from the path's origin to it, and this time value will serve as the core indicator for evaluating the node's reachability.

[0136] The real-time navigation travel time is a time length value, usually in minutes, representing the total time that a vehicle is expected to take to travel from the starting point to the destination in the real road network at the current moment.

[0137] S603. Calculate the ratio of each real-time navigation passage time to the preset baseline passage time, and use it as the time factor for each potential node.

[0138] Specifically, after obtaining the real-time navigation travel time from the starting point of the path to each potential node, these time values ​​are compared with a unified reference value—the preset baseline travel time. This preset baseline travel time is typically the theoretical shortest or average time required to travel from the starting point of the path to the potential node under ideal traffic conditions (such as during late-night smooth traffic). For each potential node, the real-time navigation travel time of that node is used as the divisor, and the preset baseline travel time is used as the divisor. The resulting ratio is the time factor corresponding to that potential node. This factor quantifies the degree of deterioration of the current real-time traffic conditions relative to ideal traffic conditions.

[0139] The preset baseline travel time is the divisor in the division operation; it is a relatively fixed reference value representing the ideal time to reach the same potential node under congestion-free conditions. The time factor is a value greater than or equal to 1 (because real-time time is usually greater than or equal to the baseline time), used to represent the inflation factor of the time cost to reach the node relative to the ideal state. A factor of 1 indicates perfect road conditions, while a larger factor indicates worse road conditions and higher time costs.

[0140] S604. Calculate the product of the time factor and the first preset weight as the time cost value corresponding to each potential node.

[0141] Specifically, after obtaining the time factor for each potential node, a pre-set parameter—the first preset weight—is read. This weight is a decimal between 0 and 1, representing the proportion of importance of travel time in the driver's decision to stop. For each potential node, the node's time factor is used as the multiplicand and multiplied by the first preset weight. The product is the time cost value corresponding to that node. This value represents the specific contribution of the time factor to the total cost after importance weighting.

[0142] The first preset weight is a globally fixed parameter that represents the relative importance of time cost in the overall cost calculation. The time cost value is a weighted numerical value, and its physical meaning is the portion of the cost caused by the time factor.

[0143] Through the above embodiments, by introducing a first preset weight, the system can flexibly adjust the tendency of the traffic acquisition strategy, incorporate preset management strategies and user preference models, making the recommendation results more intelligent and humanized, and more likely to be accepted and adopted by drivers, thereby effectively improving the success rate of traffic acquisition measures.

[0144] S605. Calculate the ratio of the real-time parking fee rate of each potential node to the preset benchmark fee rate, and use it as the fee factor corresponding to each potential node.

[0145] Specifically, when processing each potential node in a potential congestion cluster, the system reads the node's dynamically adjusted real-time parking rate and compares it with a unified reference price, i.e., a preset benchmark rate. This preset benchmark rate is typically the parking lot's regular price during off-peak hours or in non-congested conditions, or a standard reference price set by the management department. For each potential node, the system uses the node's real-time parking rate as the divisor and the preset benchmark rate as the divisor; the resulting ratio is the rate factor corresponding to that potential node. This factor quantifies the degree of fluctuation of the current real-time rate relative to the benchmark rate.

[0146] The preset base rate is a relatively fixed reference price, representing the charging standard for this node under normal conditions. The rate factor is a value that is usually greater than 0, used to represent the expansion or contraction multiple of the parking price at this node relative to the base rate. A factor of 1 indicates that the price is charged at the base rate, a factor greater than 1 indicates a premium (increase), and a factor less than 1 indicates a discount (decrease).

[0147] Through the above embodiments, by introducing a benchmark rate and calculating a rate factor, the system no longer focuses solely on the absolute value of the price. It provides a standardized indicator that purely reflects the attractiveness of the price for subsequent calculation of the comprehensive cost value. This allows price costs and time costs to be weighted and combined under a unified, dimensionless scale, thereby ensuring that the final cost ranking result can scientifically balance monetary and time expenditures and guide strategies to be more accurate and effective.

[0148] S606. Calculate the product of the rate factor and the second preset weight as the rate cost value corresponding to each potential node. The sum of the first preset weight and the second preset weight is 1.

[0149] Specifically, after obtaining the rate factor corresponding to each potential node, a pre-set parameter, namely the second pre-set weight, is read and matched with the first pre-set weight. The sum of the first and second pre-set weights is mandatory, ensuring that time and price, the two core decision factors, together constitute the entirety of the driver's decision-making model. For each potential node, the rate factor of that node is used as the multiplicand and multiplied by the second pre-set weight. The resulting product is the rate cost value corresponding to that node. This value represents the specific contribution of the price factor to the total cost after importance weighting.

[0150] The second preset weight is a globally fixed parameter that represents the relative importance of price cost in the overall cost calculation; its value complements the first preset weight. The rate cost value is a weighted figure, and its physical meaning is the portion of the cost brought about by the price factor.

[0151] S607. Calculate the sum of the time cost value and the rate cost value to obtain the comprehensive cost value corresponding to each potential node.

[0152] Specifically, after calculating the time cost and rate cost for all potential nodes, an addition operation is performed for each potential node, summing its time cost and rate cost. This means that when a driver chooses a parking lot, their decision is based on a comprehensive trade-off between time and money. These two cost values, representing different dimensions, are considered as mutually compensating components, and their summation yields a global cost metric. The result is the comprehensive cost value for that potential node. This value is a single number that comprehensively reflects the total cost (combining time and money) required to reach and use the potential node from the current route starting point (high-voltage node).

[0153] The overall cost value is a single numerical value used to comprehensively measure the total cost of going to and using a potential node. The lower the value, the higher the overall attractiveness of the node.

[0154] Through the above embodiments, the two core concerns of drivers—how long it takes to arrive and how much it costs—are integrated into a single score that can be directly compared. This provides a unique and objective basis for generating accurate, efficient, and easy-to-understand traffic diversion instructions, ensuring that the traffic diversion strategy can guide traffic flow to the overall optimal distribution in the most direct way, maximizing the utilization efficiency of urban parking resources and the overall operational efficiency of the transportation system.

[0155] S608. Sort the comprehensive cost values ​​from high to low, and select a preset number of potential nodes corresponding to the comprehensive cost values ​​from low to high, and generate a flow guidance instruction from the starting point of the path to each selected potential node.

[0156] Specifically, after obtaining the comprehensive cost values ​​of all potential nodes within a potential congestion cluster, these values ​​are sorted as a dataset. The sorting order is from low to high, meaning that the node with the lowest comprehensive cost value is at the top of the list, representing the most cost-effective option. After sorting, based on a pre-set number (e.g., top 3 or top 5), a corresponding number of potential nodes are selected starting from the top of the sorted list (i.e., the low-cost end). These selected nodes are the final destinations recommended to drivers by the system, and a traffic redirection instruction is officially generated. The core content of this instruction includes navigation path information from the fixed starting point (i.e., the high-pressure node) to each selected potential node, as well as the real-time status of these nodes (such as available parking spaces, preferential rates, etc.), forming a complete guidance solution package.

[0157] The preset quantity is a positive integer set by the system administrator. It controls the number of destination options ultimately recommended to the driver, aiming to provide multiple alternatives rather than a single choice.

[0158] Through the above embodiments, it is ensured that the final recommendation is always the optimal option under the current state. By providing a preset number of alternative options, clear guidance is given for the first choice, while also providing drivers with a certain degree of flexibility in their choices. This adapts to the preferences of different drivers, improves the compliance rate of instructions, and can proactively, intelligently, and efficiently guide traffic flow from congestion hotspots to resource-rich areas. Thus, at the operational level, it ultimately achieves the optimized allocation of city-level parking resources and the coordinated management of regional traffic congestion.

[0159] The city-level parking resource optimization system in the embodiments of this invention is described below from the perspective of hardware processing. Please refer to [link / reference]. Figure 3 , Figure 3 This is a schematic diagram of a city-level parking resource optimization system provided in an embodiment of this application.

[0160] It should be noted that, Figure 3 The structure of the city-level parking resource optimization system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0161] The city-level parking resource optimization system in the embodiments of this invention is described below from the perspective of hardware processing. Please refer to [link / reference]. Figure 3 , Figure 3 This is a schematic diagram of the structure of a city-level parking resource optimization system provided in an embodiment of this application.

[0162] It should be noted that, Figure 3 The structure of the city-level parking resource optimization system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0163] like Figure 3 As shown, the city-level parking resource optimization system includes a central processing unit 701, which can perform various appropriate actions and processes based on a program stored in a read-only memory 702 or a program loaded from a storage section 708 into a random access memory 703, such as executing the methods described in the above embodiments. The random access memory 703 also stores various programs and data required for system operation. The central processing unit 701, the read-only memory 702, and the random access memory 703 are interconnected via a bus 704. An input / output interface 705 is also connected to the bus 704.

[0164] The following components are connected to the input / output interface 705: an input section 706 including audio input devices, push-button switches, etc.; an output section 707 including a liquid crystal display (LCD) and audio output devices, indicator lights, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN (Local Area Network) card, modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the input / output interface 705 as needed. A removable medium 711, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 710 as needed so that computer programs read from it can be installed into the storage section 708 as needed.

[0165] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing a computer program for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 709, and / or installed from removable medium 711. When the computer program is executed by central processing unit 701, it performs the various functions defined in the present invention. It should be noted that specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0166] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0167] Specifically, the city-level parking resource optimization system of this embodiment includes a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, it implements the city-level parking resource optimization method provided in the above embodiment.

[0168] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the city-level parking resource optimization system described in the above embodiments; or it may exist independently and not assembled into the city-level parking resource optimization system. The storage medium carries one or more computer programs, which, when executed by a processor of the city-level parking resource optimization system, cause the city-level parking resource optimization system to implement the city-level parking resource optimization method provided in the above embodiments.

[0169] The above description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of other embodiments of this disclosure upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A city-level parking resource optimization method, characterized in that, include: Obtain real-time operational data and geographic location of all parking lots within the target city area; Construct an urban parking topology graph with each parking lot as a node and the connecting roads of each node as edges; Based on the real-time traffic data of the target city area, the real-time weights of each edge are obtained; Based on the real-time operational data, the real-time parking pressure value of each node is calculated, wherein the real-time operational data includes at least the static capacity and real-time occupancy rate of the corresponding parking lot; Nodes whose real-time parking pressure value is greater than or equal to the first preset pressure value are identified as high-pressure nodes, and nodes whose real-time parking pressure value is less than the first preset pressure value are identified as potential nodes. Based on the real-time weights, pressure transmission simulation is performed on each high-pressure node in the urban parking topology map to obtain the simulated parking pressure value of each node. The potential nodes whose simulated parking pressure value is greater than the second preset pressure value and less than the first preset pressure value, together with a corresponding high-pressure node, are identified as a potential congestion cluster. The second preset pressure value is less than the first preset pressure value, and the actual traffic distance between the potential node and the corresponding high-pressure node is less than a preset distance threshold. For each of the aforementioned potential congested clusters, the following steps are performed to optimize resources: Based on the simulated parking pressure value corresponding to each target node, calculate the price adjustment coefficient corresponding to each target node, where the target node is any node of the potential congestion cluster; Based on the price adjustment coefficient, the real-time parking fee rate after price adjustment for each target node is obtained; Based on the real-time parking rates mentioned above, a referral instruction is generated.

2. The method according to claim 1, characterized in that, The real-time weights of each edge are obtained based on the real-time traffic data of the target city area, including: Obtain the actual traffic distance and preset traffic speed corresponding to each edge; Calculate the ratio of each actual traffic distance to the preset travel speed to obtain the simulated travel time corresponding to each edge; Calculate the reciprocal of each simulated travel time to obtain the real-time weight corresponding to each edge.

3. The method according to claim 1, characterized in that, The real-time operational data also includes the entry count and exit count of the corresponding node within a preset time window. The preset time window is a sliding time window ending at the current moment. The calculation of the real-time parking pressure value for each node based on the real-time operational data includes: Calculate the difference between the inlet count and the outlet count corresponding to each node within the preset time window, and use it as the net inflow count of each node within the preset time window; Calculate the ratio of each net inflow count to each static capacity to obtain the net inflow rate of each node in the preset time window; Calculate the average of each net inflow rate and each real-time occupancy rate to obtain the real-time parking pressure value corresponding to each node.

4. The method according to claim 1, characterized in that, Based on the real-time weights, pressure transmission simulation is performed on each high-pressure node in the urban parking topology map to obtain the simulated parking pressure value of each node, including: In each conduction simulation, the following steps are performed for each of the high-voltage nodes: Obtain each of the potential nodes directly adjacent to the high-voltage node; The real-time weights corresponding to the edges connecting the high-voltage node and each of the directly adjacent potential nodes are used as the distance attenuation coefficients corresponding to each of the directly adjacent potential nodes. Calculate the product of each distance attenuation coefficient and the real-time parking pressure value of the high-pressure node to obtain the single transmission pressure value transmitted to the corresponding directly adjacent potential node; The single transmission pressure value received by each potential node is accumulated and added to the pressure cache value of the corresponding potential node; When the conduction simulation is completed a preset number of times, for each potential node, the corresponding pressure cache value is used as the total conduction pressure value of the corresponding potential node; Based on the total transmission pressure value, the simulated parking pressure value of each node is obtained.

5. The method according to claim 4, characterized in that, The process of obtaining the simulated parking pressure value for each node based on the total transmitted pressure value includes: For each potential node, the corresponding real-time parking pressure value is added to the corresponding total transmission pressure value to obtain the simulated parking pressure value of the corresponding potential node. For each high-voltage node, the corresponding real-time parking pressure value is used as the simulated parking pressure value of the corresponding high-voltage node.

6. The method according to claim 1, characterized in that, The calculation of the price adjustment coefficient corresponding to each target node based on the simulated parking pressure value corresponding to each target node includes: Obtain the maximum and minimum pressure values ​​among the simulated parking pressure values ​​in the potential congestion cluster; Calculate the difference between the maximum pressure value and the minimum pressure value, and use it as the first difference value; For each target node, the following steps are performed to obtain the corresponding price adjustment coefficient: Calculate the difference between the simulated parking pressure value and the minimum pressure value, and use it as the second difference value; The initial price adjustment coefficient of the target node is obtained by calculating the ratio of the second difference to the first difference; If the initial price adjustment coefficient is less than the preset lower limit, then the price adjustment coefficient is set to the preset lower limit. If the initial price adjustment coefficient is greater than the preset upper limit value, then the price adjustment coefficient is set to the preset upper limit value; If the initial price adjustment coefficient is greater than or equal to the preset lower limit and less than or equal to the preset upper limit, then the initial price adjustment coefficient is used as the price adjustment coefficient of the target node.

7. The method according to claim 1, characterized in that, The process of generating referral instructions based on the aforementioned real-time parking rates includes: The high-voltage node in the potential congestion cluster is used as the starting point of the diversion path; Obtain the real-time navigation passage time between the starting point of the path and each potential node in the potential congestion cluster; Calculate the ratio of each real-time navigation travel time to the preset baseline travel time, and use it as the time factor corresponding to each potential node; Calculate the product of the time factor and the first preset weight, and use it as the time cost value corresponding to each potential node; Calculate the ratio of the real-time parking fee rate of each potential node to the preset benchmark fee rate, and use it as the fee factor corresponding to each potential node; The product of the rate factor and the second preset weight is calculated as the rate cost value corresponding to each potential node, and the sum of the first preset weight and the second preset weight is 1. Calculate the sum of the time cost value and the rate cost value to obtain the comprehensive cost value corresponding to each potential node; The comprehensive cost values ​​are sorted from high to low, and a preset number of potential nodes corresponding to the comprehensive cost values ​​are selected from low to high to generate the traffic guidance instructions from the starting point of the path to each of the selected potential nodes.

8. A city-level parking resource optimization system, characterized in that, The city-level parking resource optimization system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the city-level parking resource optimization system to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on the city-level parking resource optimization system, the city-level parking resource optimization system performs the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on a city-level parking resource optimization system, the city-level parking resource optimization system performs the method as described in any one of claims 1-7.

Images