Parking guiding method and system based on data analysis

Abstract

The invention relates to a parking guidance method and system based on data analysis, and the method comprises the steps: obtaining parking space data of a parking lot, carrying out the region division, and obtaining the corresponding region parking space information; real-time parking demand information of a user is obtained, parking space association is carried out on the parking demand information and the regional parking space information, and a corresponding real-time parking group is obtained; historical parking data of the parking lot is acquired for parking demand prediction, and a corresponding parking demand trend is obtained; performing parking planning analysis on the real-time parking group and the parking demand trend to obtain initial parking space distribution information; acquiring real-time parking demand information of the parking lot, and performing parking prediction according to the parking demand trend to obtain a corresponding parking guide scheme; and performing parking space guiding analysis on the initial parking space distribution information and the parking guiding scheme to obtain global parking guiding information. According to the invention, the parking resource distribution conditions of different areas can be evaluated more accurately.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile parking, and in particular to a parking guidance method and system based on data analysis. Background Art

[0002] With the continuous advancement of urbanization and the continuous growth of the number of motor vehicles, the problem of parking difficulties has become increasingly prominent, and intelligent parking management has become an important issue in urban governance. At present, although large parking lots are equipped with basic management systems, these systems can often only provide simple parking space occupancy status information, lacking in-depth analysis of parking demand and accurate guidance capabilities. Existing parking guidance solutions usually only consider real-time parking space vacancy, ignoring multi-dimensional factors such as regional distribution characteristics, historical data patterns, and future demand trends. This single guidance method is difficult to achieve optimal allocation of parking resources, and it is easy to cause an imbalance in which parking spaces in some areas are idle for a long time while other areas are congested. How to make full use of data analysis technology to build a parking management system that can comprehensively consider multi-dimensional information and provide intelligent guidance solutions has become a problem that needs to be solved urgently. Summary of the invention

[0003] The main purpose of the present invention is to provide a parking guidance method and system based on data analysis, which can more accurately evaluate the distribution of parking resources in different areas, thereby improving the accuracy of parking guidance.

[0004] To achieve the above object, the present invention provides a parking guidance method based on data analysis, comprising: Obtain parking space data of the parking lot for regional division and obtain corresponding regional parking space information; Acquire the user's real-time parking demand information, associate the parking demand information with the regional parking space information, and obtain a corresponding real-time parking group; Acquire historical parking data of the parking lot to perform parking demand prediction and obtain corresponding parking demand trends; Performing parking planning analysis on the real-time parking group and the parking demand trend to obtain initial parking space allocation information; Acquire the real-time parking demand information of the parking lot, perform parking prediction based on the parking demand trend, and obtain a corresponding parking guidance plan; The initial parking space allocation information and the parking guidance plan are subjected to parking space guidance analysis to obtain global parking guidance information.

[0005] Furthermore, the acquisition of parking space data of the parking lot is divided into regions to obtain corresponding regional parking space information, including: Performing spatial grid division on the parking space data to obtain initial grid unit data; Performing multi-layer grid index construction on the initial grid unit data to obtain a hierarchical grid index structure; Calculating the grid parking space density on the hierarchical grid index to obtain grid unit parking space density distribution data; Clustering adjacent grids according to the grid unit parking space density distribution data to obtain an initial area division result; Perform parking space coding mapping on the initial area division result to obtain parking space coding information in the area; Data matching is performed according to the parking space coding information in the area to obtain the parking space information in the area.

[0006] Furthermore, the acquiring of the user's real-time parking demand information, associating the parking demand information with the regional parking space information, and obtaining a corresponding real-time parking group includes: Extract information from parking requests submitted by users to obtain real-time parking demand information; Performing area matching screening according to the real-time parking demand information to obtain a set of candidate parking areas; Performing parking space availability analysis on the candidate parking area set to obtain real-time vacant parking space data; Calculate the distance of each candidate area according to the real-time vacant parking space data to obtain the shortest path data; Prioritizing the shortest path data and the real-time vacant parking space data to obtain an initial parking space allocation sequence; Performing time conflict detection on the initial parking space allocation sequence to obtain a valid parking space combination; Demand grouping is performed according to the effective parking space combination to obtain the real-time parking group.

[0007] Furthermore, the acquiring of historical parking data of the parking lot to perform parking demand prediction and obtain a corresponding parking demand trend includes: Performing sequence decomposition on the historical parking data to obtain a parking feature sequence; Perform parking periodicity analysis according to the parking feature sequence to obtain time distribution characteristics; Performing clustering processing on the time distribution characteristics to obtain parking regularity characteristics; Performing parking correlation analysis according to the parking regularity characteristics to obtain location correlation characteristics; Performing demand forecasting on the location-related features to obtain parking demand change rules; The parking demand trend is obtained by performing trend analysis on the parking feature sequence according to the parking demand change rule.

[0008] Furthermore, performing parking planning analysis on the real-time parking group and the parking demand trend to obtain initial parking space allocation information includes: Extracting vehicle information features from the real-time parking group to obtain vehicle travel pattern data; Performing entropy calculation processing on the parking demand trend according to the vehicle travel pattern data to obtain parking entropy distribution data; Performing bidirectional dynamic propagation prediction on the parking entropy value distribution data to obtain dynamic prediction data of parking spaces; Performing hierarchical allocation of parking space resources according to the dynamic prediction data of parking spaces to obtain hierarchical configuration weight data; Performing multi-constraint group intelligent optimization on the hierarchical configuration weight data to obtain parking space optimization configuration data; A global dynamic balance calculation is performed on the parking space optimization configuration data to obtain the initial parking space allocation information.

[0009] Furthermore, the real-time parking demand information of the parking lot is obtained, and parking prediction is performed based on the parking demand trend to obtain a corresponding parking guidance plan, including: Performing periodic decomposition on the parking demand information to obtain a segmented feature sequence; Calculating the parking demand trend index according to the segmented feature sequence to obtain a parking correlation index; Deeply integrating the parking correlation indicators to obtain a regional parking weight matrix; Calculate the parking demand probability distribution according to the regional parking weight matrix to obtain the regional parking density distribution; Clustering the parking density distribution in the area to obtain the parking regularity characteristics of the area; Predicting parking flow according to the parking pattern characteristics of the area to obtain dynamic parking prediction data; Intelligently configure parking resources based on the dynamic parking prediction data to obtain a parking guidance plan.

[0010] Furthermore, the performing parking guidance analysis on the initial parking space allocation information and the parking guidance scheme to obtain global parking guidance information includes: Performing density distribution calculation on the initial parking space allocation information to obtain a parking space space distribution heat map; Perform multi-objective optimization on the parking guidance scheme according to the parking space space distribution heat map to obtain a guidance strategy; Performing multi-source shortest path calculation on the guidance strategy to obtain a dynamic obstacle avoidance guidance path; Performing traffic flow entropy analysis based on the dynamic obstacle avoidance guidance path to obtain a path load index; Performing joint iterative optimization on the path load index and the initial parking space allocation information to obtain an adaptive scheduling solution; Regional balance factor mapping is performed according to the adaptive scheduling scheme to obtain global parking guidance information.

[0011] The present invention further provides a parking guidance system based on data analysis, which is applied to any one of the above-mentioned parking guidance methods based on data analysis, comprising: A collection module, which is used to obtain parking space data of the parking lot, divide the area into regions, and obtain corresponding regional parking space information; An analysis module, the analysis module is used to obtain the user's real-time parking demand information, associate the parking demand information with the regional parking space information, and obtain a corresponding real-time parking group; A correlation module, the correlation module is used to obtain historical parking data of the parking lot to perform parking demand prediction and obtain a corresponding parking demand trend; A processing module, the processing module is used to perform parking planning analysis on the real-time parking group and the parking demand trend to obtain initial parking space allocation information; A control module, the control module is used to obtain real-time parking demand information of the parking lot, perform parking prediction based on the parking demand trend, and obtain a corresponding parking guidance plan; An execution module is used to perform parking guidance analysis on the initial parking space allocation information and the parking guidance plan to obtain global parking guidance information.

[0012] The present invention provides a parking guidance method and system based on data analysis, which has the following beneficial effects: By dividing the parking lot data into regions and analyzing parking spaces, combined with real-time user demand information, the distribution of parking resources in different regions can be more accurately evaluated, thereby improving the accuracy of parking guidance and providing a more scientific basis for parking space allocation. By correlating user demand data with regional parking space information, refined management of parking resources in different regions can be achieved, which helps to achieve reasonable allocation and avoid resource waste. Trend analysis of parking demand based on historical data can ensure that the system can operate efficiently in different time periods and scenarios, reduce vacant parking spaces, and improve the overall utilization rate of parking lots. By comprehensively analyzing real-time parking groups and demand trends, a more reasonable parking guidance plan can be formulated, and the optimization operation of the entire parking lot can be achieved through global adjustment strategies, thereby effectively utilizing parking resources and reducing congestion and queuing. And by considering the impact of real-time demand on parking guidance, the parking space allocation strategy can be flexibly adjusted according to the characteristics of different time periods and demand changes, making the system more adaptable to dynamically changing parking scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 is a flow chart of a parking guidance method based on data analysis provided by the present invention; Figure 2 It is a structural diagram of a parking guidance system based on data analysis provided by the present invention.

[0014] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0016] The present invention is further described below in conjunction with the accompanying drawings and specific implementation methods.

[0017] Reference Figure 1 As shown, the present invention provides 1. A parking guidance method based on data analysis, characterized in that it includes: Step S1: Obtain parking space data of the parking lot and divide it into regions to obtain corresponding regional parking space information; Step S2: obtaining the user's real-time parking demand information, associating the parking demand information with the regional parking space information, and obtaining the corresponding real-time parking group; Step S3: Obtain historical parking data of the parking lot to predict parking demand and obtain corresponding parking demand trends; Step S4: Perform parking planning analysis on the real-time parking groups and parking demand trends to obtain initial parking space allocation information; Step S5: Acquire the real-time parking demand information of the parking lot, make parking predictions based on the parking demand trend, and obtain corresponding parking guidance plans; Step S6: Perform parking guidance analysis on the initial parking space allocation information and the parking guidance plan to obtain global parking guidance information.

[0018] Based on the above steps, the detailed process is as follows: Step S1: Get the floor plan of the parking lot, including the specific location coordinates, parking space dimensions and parking space numbers of all parking spaces. Reasonable regional division is carried out according to the physical characteristics of the parking lot, such as building pillars, passages, entrance and exit locations, etc. Regional division uses a grid method or a clustering algorithm to divide parking spaces with similar geographical locations and similar characteristics into the same area. Each area needs to have a unique regional identifier and record information such as the total number of parking spaces in the area, the current number of vacant parking spaces, and the coordinates of the center point of the area. Establish an adjacency matrix between regions to record the connectivity and distance information between regions, which is of great significance for subsequent parking space allocation and guidance. After the regional division is completed, it is necessary to establish a parking space information table in the database to store the basic attributes of each parking space, such as parking space type (standard, compact, barrier-free, etc.), area to which it belongs, and usage status. This information will provide basic data support for subsequent real-time parking space allocation.

[0019] Step S2: Receive parking demands from users through mobile applications or vehicle terminals, including information such as the user's current location, estimated arrival time, vehicle type, and estimated parking duration. The system will establish an association between user demands and available parking spaces based on these demand information and the parking space information in the step area. In the association process, multiple factors need to be considered: first, the availability of parking spaces, screening out currently vacant parking spaces; second, location adaptability, calculating the driving distance and time to each available parking space based on the user's current location and estimated arrival time; and finally, user preference matching, such as whether a charging pile is needed, whether it is close to an elevator, and other special needs. These matching parking spaces are organized into real-time parking groups, each of which contains multiple candidate parking spaces and is sorted according to the comprehensive score. This score takes into account multiple dimensions such as distance, convenience, and special needs matching.

[0020] Step S3: Collect and organize historical parking data, including information such as parking space occupancy rate, parking duration distribution, and usage frequency in different areas in each time period. Through time series analysis methods such as moving average, exponential smoothing, or more complex ARIMA models, the periodic patterns of parking demand are identified, such as the difference between weekdays and weekends, the characteristics of morning and evening peaks, etc. At the same time, it is also necessary to consider the impact of external factors on parking demand, such as weather conditions, holidays, surrounding activities, etc. The system uses machine learning algorithms such as random forests or neural networks to take these features as input and train the prediction model. The prediction time span can be short-term (next few hours) to medium-term (next few days), and the prediction results should include demand estimates for different time periods and different areas. These prediction results will form a parking demand trend chart, which will provide an important reference for subsequent parking planning.

[0021] Step S4: Based on the two main goals of the overall operational efficiency of the parking lot and user satisfaction. For each real-time parking group, evaluate the impact of the allocation scheme of the candidate parking spaces on future parking demand. For example, if the forecast shows that a certain area will have a high parking demand in the next hour, appropriately reduce the allocation weight of the currently available parking spaces in the area to reserve a certain number of parking spaces to cope with the peak period. The parking space turnover efficiency also needs to be considered in the optimization process. According to the expected parking time declared by the user, long-term parking needs are directed to non-hot spots, while short-term parking needs are arranged in convenient locations. Use a heuristic algorithm (such as a genetic algorithm or an ant colony algorithm) to solve this optimization problem and obtain an optimal parking space allocation plan at the current moment. This plan not only meets the parking needs of current users, but also takes into account the overall operational efficiency of the parking lot.

[0022] Step S5: Combine real-time data with the prediction model to generate a dynamic parking guidance plan. The system collects dynamic information of the parking lot in real time through various sensors and monitoring equipment, including vehicle entry flow, current parking space occupancy status, vehicle driving trajectory, etc. These real-time data will be compared and calibrated with the prediction model. If there is a significant deviation between the actual situation and the prediction, the prediction parameters will be adjusted quickly. Based on the updated prediction results, a short-term (15-30 minutes) parking lot status prediction is generated, including the congestion level of each area, possible parking space shortage, etc. The system will also consider dynamic factors in the parking lot, such as temporarily closed lanes, parking spaces under maintenance, etc., and adjust feasible driving routes in real time. Finally, the system will plan an optimal driving route for each incoming vehicle. This route not only considers the distance to the target parking space, but also avoids possible congested areas to ensure smooth driving.

[0023] Step S6: Match and adjust the initial parking space allocation plan with the dynamic guidance suggestions. If it is found that some of the original allocation plans may cause local congestion or affect the passage of other vehicles at the current moment, adjust the allocation plan in real time. Global parking guidance information includes multiple levels: one is the parking space level, which clearly allocates a specific target parking space for each user; the second is the path level, which provides the optimal driving route from the entrance to the target parking space; the third is the time level, which estimates the expected time to reach the target parking space. This information will be displayed to users through various channels, such as mobile phone APP, in-vehicle navigation, parking lot LED display, etc. Establish a feedback mechanism to collect user experience data during the actual parking process for continuous optimization of the guidance algorithm. When an emergency occurs, such as a parking space being occupied or a road being temporarily diverted, the system can quickly recalculate and provide alternative solutions.

[0024] The present invention provides a parking guidance method and system based on data analysis, which has the following beneficial effects: By dividing the parking lot data into regions and analyzing parking spaces, combined with real-time user demand information, the distribution of parking resources in different regions can be more accurately evaluated, thereby improving the accuracy of parking guidance and providing a more scientific basis for parking space allocation. By correlating user demand data with regional parking space information, refined management of parking resources in different regions can be achieved, which helps to achieve reasonable allocation and avoid resource waste. Trend analysis of parking demand based on historical data can ensure that the system can operate efficiently in different time periods and scenarios, reduce vacant parking spaces, and improve the overall utilization rate of parking lots. By comprehensively analyzing real-time parking groups and demand trends, a more reasonable parking guidance plan can be formulated, and the optimization operation of the entire parking lot can be achieved through global adjustment strategies, thereby effectively utilizing parking resources and reducing congestion and queuing. And by considering the impact of real-time demand on parking guidance, the parking space allocation strategy can be flexibly adjusted according to the characteristics of different time periods and demand changes, making the system more adaptable to dynamically changing parking scenarios.

[0025] In one embodiment, the parking space data of the parking lot is obtained to divide the area into regions, and the corresponding regional parking space information is obtained, including: The parking space data is used to divide the area, so as to obtain the corresponding regional parking space information. When dividing the space grid, the parking space data of the parking lot is mapped to the two-dimensional coordinate system, and the entire parking lot is divided into multiple regular grid cells according to the preset grid size (such as 10 meters × 10 meters). The Cartesian coordinate system is used in the division process, with the lower left corner of the parking lot as the origin (0,0), the right as the positive direction of the x-axis, and the upward as the positive direction of the y-axis. Each grid cell records the coordinate values ​​of the four vertices and the position coordinate information of all the parking spaces in its area to obtain the initial grid cell data. For parking spaces that cross the grid boundary, they are divided into grid cells with a larger area.

[0026] In the process of constructing a multi-layer grid index, a quadtree structure is used to manage the initial grid units in layers. Starting from the top layer, the entire parking lot is taken as the root node. When the number of parking spaces in the node exceeds the preset threshold (such as 50) or the area is greater than the set value (such as 400 square meters), the node is divided into four sub-grids. The segmentation process is performed recursively until the number of parking spaces in the grid is lower than the threshold or the area is less than the minimum limit. Each grid node stores its level number, range coordinates, the included parking space information and child node pointers to obtain a hierarchical grid index structure. The index structure realizes fast spatial query and regional positioning through the parent-child node relationship.

[0027] The lowest level grid cells are processed in the grid parking density calculation stage. The density value calculation formula is ρ=N / S, where N is the number of parking spaces in the grid and S is the actual available area of ​​the grid (excluding unavailable spaces such as walls and columns). In order to eliminate the influence of local outliers, Gaussian smoothing is introduced to perform neighborhood weighted average on the density value of each grid. The smoothed density distribution can better reflect the overall characteristics of the region, which is conducive to subsequent clustering analysis.

[0028] The clustering process of adjacent grids is based on the improved DBSCAN algorithm. Set the density difference threshold ε (such as 0.2) and the minimum cluster area Smin (such as 200 square meters). Clustering starts from the highest density grid, and the adjacent grids with a density difference less than ε and shared boundaries are classified into the same area. The regularity of the regional shape is also considered in the clustering process to avoid the generation of too narrow or irregular regional shapes. For scattered grids with too low density, they are merged into the adjacent area according to the principle of proximity.

[0029] The parking space coding mapping adopts a hierarchical coding system. The area number uses two digits (01-99) and is numbered from north to south and from west to east. The parking space number is added with a three-digit serial number (001-999) after the area number and is coded in relative position order. The coding rules also take into account the actual navigation needs, so that adjacent parking spaces have continuous or similar numbers. For special parking spaces (such as charging piles and barrier-free parking spaces), a type identifier is added to the number.

[0030] The data matching phase builds a complete regional parking space information database. The database contains basic regional information (number, range, entrance and exit locations), parking space details (number, type, coordinates, orientation angle, occupancy status) and statistical information (total number of parking spaces, number of remaining parking spaces, utilization rate). The database implements bidirectional indexing of regional information and parking space information, and supports the functions of querying parking spaces by region and locating regions by parking spaces. The database regularly updates parking space status information to ensure data timeliness, and saves historical data for utilization rate analysis and prediction. Abnormal data is marked and processed during the matching process to ensure data consistency and reliability.

[0031] The parking guidance method based on data analysis in this embodiment has significant technical effects: the method realizes accurate division and rapid retrieval of parking space by combining spatial grid division with quadtree index structure, effectively improving the accuracy of regional division and data processing efficiency. The density-based grid clustering algorithm is adopted, combined with Gaussian smoothing processing, to eliminate the interference of local outliers, making regional division more reasonable and more adaptable. The design of the hierarchical coding system not only meets the actual navigation needs, but also facilitates system maintenance and expansion, and improves the accuracy of parking guidance. The complete database design supports bidirectional indexing function, realizes efficient association between regional information and parking space information, and ensures the real-time and reliability of data through regular updates and exception handling mechanisms.

[0032] In one embodiment, the real-time parking demand information of the user is obtained, and the parking demand information is associated with the parking space information of the area to obtain the corresponding real-time parking group, including: During the extraction and processing of parking request information, the parking request form submitted by the user is parsed, including the extraction of key information such as parking period, parking duration, and destination location, and a standardized parking demand information data structure is formed through standardization. The original request data is deconstructed in JSON format to extract the information of each field filled in by the user. The extracted information is converted into data type and format normalized, and the time information is uniformly converted into timestamp format, and the location information is uniformly converted into latitude and longitude coordinate format. After processing, real-time parking demand information is obtained, which contains fields such as user ID, license plate number, estimated arrival time, estimated parking duration, destination latitude and longitude coordinates, vehicle type, special demand mark, etc.

[0033] In the area matching screening stage, based on the user's destination location information, the search radius is set to 1.5 kilometers, and all parking areas are retrieved within this range. The screening rules for parking areas include three dimensions: area type matching, area opening period matching, and parking space specification matching. The area type matching judgment rule is: roadside parking space> outdoor parking lot> indoor parking garage; the area opening period matching judgment rule is: the area whose opening period completely covers the user's demand period has the highest priority; the parking space specification matching judgment rule is: the parking space size meets the user's vehicle model requirements. The parking areas within the search range are evaluated for rule matching one by one, and the areas with scores that reach the threshold will be included in the candidate parking area set, which contains information such as area ID, area type, area location, total number of parking spaces, and area score.

[0034] In the parking space availability analysis phase, real-time parking space status detection is performed on each parking area in the candidate area set. The detection content includes information such as the total number of parking spaces in the area, the number of occupied parking spaces, the number of reserved parking spaces, and the number of temporarily locked parking spaces. The real-time data collected by the parking space detection equipment is combined with the historical reservation data in the next 4 hours, and the time series prediction model is applied to calculate the estimated number of vacant parking spaces in each area during the user's target time period. The prediction model considers multiple variables such as periodic laws, the impact of temporary events, and weather factors. The analysis results form real-time vacant parking space data, and the data structure contains fields such as area ID, current vacant number, predicted vacant number, and available parking space location.

[0035] In the shortest path calculation process, the improved Dijkstra algorithm is used to calculate the walking distance from each candidate parking area to the destination with the user's destination as the end point. When planning the path, environmental factors such as road conditions, sidewalk width, number of traffic lights, slope, and sunshade facilities are comprehensively considered to generate the optimal walking route. The algorithm assigns different weights to each influencing factor and obtains a comprehensive path score through weighted calculation. The path calculation result forms the shortest path data, which includes information such as the starting point coordinates, the end point coordinates, the path length, the estimated walking time, and the path score.

[0036] In the priority sorting process, a multi-dimensional comprehensive evaluation is conducted on the shortest path data and real-time vacant parking space data. The evaluation indicators include walking distance, parking space adequacy, price level, and historical use evaluation. The weight distribution of each dimension is: walking distance 0.4, parking space adequacy 0.3, price level 0.2, and historical evaluation 0.1. The comprehensive score of each candidate area is obtained through weighted calculation, and the scores are sorted from high to low to form an initial parking space allocation sequence, which contains detailed information such as area sorting, comprehensive score, and scores of each dimension.

[0037] In the time conflict detection phase, the parking spaces in the initial allocation sequence are analyzed for conflicts in the time dimension. The detection rules include: reservation period overlap detection, buffer time requirement detection, and peak capacity control detection. The reservation period overlap shall not exceed 80%, and a 15-minute buffer time shall be reserved between adjacent reservations. The peak parking space reservation rate shall not be less than 10%. By sliding the time window, the reservation situation of each time period is statistically analyzed to identify time conflict points. After conflict detection and screening, an effective parking space combination is obtained, which ensures feasibility in the time dimension.

[0038] In the demand grouping processing stage, demands with similar parking characteristics are clustered and integrated. Clustering characteristics include three dimensions: time similarity, space similarity, and user attribute similarity. The time similarity criterion is that the overlap of reservation time periods exceeds 60%; the space similarity criterion is that the straight-line distance between destinations is less than 500 meters; the user attribute similarity criterion is the matching degree of user type, vehicle type, and charging standard. A hierarchical clustering algorithm is used to integrate similar demands into real-time parking groups. Each parking group contains information such as group ID, member list, common characteristics, priority, etc., which provides a basis for subsequent batch processing.

[0039] This embodiment achieves accurate matching of parking demand and parking space resources by adopting a parking guidance method based on multi-dimensional data analysis and intelligent matching. Parking requests are processed based on standardized data structures to ensure the accuracy and completeness of information extraction. The accuracy of parking area screening is improved by setting a reasonable search radius and multi-dimensional matching rules. The availability of parking spaces is analyzed using a time series prediction model, and the accuracy of parking space allocation is effectively improved by combining real-time detection data and historical data. An improved path planning algorithm and a multi-factor comprehensive evaluation method are used to achieve intelligent recommendation of the optimal parking solution. Through strict time conflict detection and demand clustering analysis, resource allocation conflicts are avoided and parking space utilization efficiency is improved. This method not only optimizes the allocation efficiency of parking resources, but also improves the user's parking experience, providing effective technical support for smart parking management.

[0040] In one embodiment, historical parking data of a parking lot is obtained to predict parking demand and obtain corresponding parking demand trends, including: By collecting the historical parking data of the parking lot as the basic data set, the data set is subjected to sequence decomposition processing. The sequence decomposition process uses the time series decomposition method to decompose the original parking data into three components: trend item, seasonal item and random item. The trend item reflects the long-term trend of parking volume, which is reflected in the overall trend of parking lot utilization rate; the seasonal item reflects the periodic change characteristics of parking behavior, reflecting the parking mode of fixed periods such as weekdays and weekends; the random item contains irregular fluctuation components, representing temporary and occasional changes in parking demand. After decomposition processing, the parking feature sequence is obtained, which contains core feature indicators such as parking time distribution, parking space turnover rate, and peak hour occupancy rate.

[0041] The decomposed parking feature sequence is subjected to periodic analysis, and the Fourier transform method is used to identify the main period in the time series. The periodic characteristics of the time series are determined by calculating the autocorrelation coefficient and the partial autocorrelation coefficient, where the autocorrelation coefficient threshold is set to ±0.5 and the partial autocorrelation coefficient threshold is set to ±0.3. Based on the periodic analysis results, the distribution characteristics of parking behavior on different time scales are extracted, including intraday distribution characteristics (morning peak 7:00-9:00, evening peak 17:00-19:00), intraweek distribution characteristics (weekday occupancy rate 85%, weekend occupancy rate 60%) and seasonal distribution characteristics (peak season months July-September, off-season months January-March), forming a complete time distribution feature set.

[0042] The time distribution characteristics are clustered, and the K-means clustering algorithm is used to classify similar time distribution patterns into one category. The number of clusters is set to 3-5 during the clustering process, and the silhouette coefficient is used to evaluate the clustering effect. The silhouette coefficient threshold is set to 0.6. Euclidean distance is used as the similarity metric in cluster analysis, and the upper limit of the number of iterations is set to 100 times. The clustering results reflect the parking rules in different time periods, such as the peak period on weekdays (occupancy rate>90%), the off-peak period on weekdays (occupancy rate 60%-90%), and the weekend leisure period (occupancy rate 40%-60%), which constitute the parking regularity characteristics.

[0043] Based on the characteristics of parking rules, spatial correlation analysis is performed to calculate the correlation coefficient of parking volume between different regions. The Pearson correlation coefficient is used for correlation analysis, and the threshold of the determination coefficient is set to 0.6, and the confidence level is 95%. By analyzing the changing relationship of parking volume in different regions, regional combinations with linkage effects are identified, such as commercial areas and residential areas, office areas and commercial areas, etc. In the process of spatial correlation analysis, the geographic weight matrix is ​​introduced to consider the distance attenuation effect between regions, and the weight calculation adopts the inverse distance weighting method. Finally, the location correlation characteristics reflecting the law of parking space distribution are obtained, including the parking spillover effect between regions, the law of parking demand transfer, etc.

[0044] The location-related features are combined with the time series forecasting model to forecast demand. The forecasting model uses the ARIMA model. The model parameters (p, d, q) are optimized and selected according to the AIC criterion, and the AIC threshold is set to -500. The model training adopts the rolling time window method, with the window length set to 90 days and the step length set to 1 day. The forecasting process takes into account the influence of external factors such as holidays, weather, and surrounding activities, establishes a multivariate regression relationship, and the goodness of fit R² of the regression model is required to be greater than 0.8. Generate parking demand forecast results for future time periods. The forecast time is divided into short-term (within 24 hours), medium-term (within 1 week) and long-term (within 1 month). The forecast accuracy requirements are 90%, 85% and 80% respectively, forming a law of parking demand changes.

[0045] The time series decomposition-reconstruction method is used to compare and analyze the predicted demand change pattern with the original parking feature sequence. The trend similarity and change rate are calculated to determine the change trend of parking demand. The trend similarity is calculated using the DTW (dynamic time warping) algorithm, and the similarity threshold is set to 0.8; the change rate is obtained by calculating the mean of the first-order difference sequence, which is used to determine the speed of demand growth or decline. The trend analysis results are used to guide the operation and management of parking lots and achieve accurate parking demand prediction and guidance. The reliability of the prediction results is evaluated by the root mean square error (RMSE) and the mean absolute percentage error (MAPE), with the RMSE threshold set to 10 and the MAPE threshold set to 15%.

[0046] This embodiment uses the time series decomposition method to process historical parking data, realizes the accurate extraction of parking features, and effectively identifies core indicators such as parking lot utilization rate and parking space turnover rate. Based on the periodic analysis of Fourier transform and K-means clustering algorithm, the time distribution law of parking behavior is accurately grasped, providing data support for parking management in different time periods. The introduction of geographic weight matrix for spatial correlation analysis effectively identifies the parking spillover effect and demand transfer law between regions, and improves the utilization efficiency of parking resources. The ARIMA model is used in combination with external factors to predict demand, achieving a prediction accuracy of 90% in the short term, 85% in the medium term and 80% in the long term, greatly improving the accuracy of parking demand prediction. Trend analysis is performed through the time series decomposition-reconstruction method, and a complete prediction and evaluation mechanism is established to ensure the reliability of the parking guidance plan, providing strong support for the intelligent operation and management of parking lots.

[0047] 5. The parking guidance method based on data analysis according to claim 1 is characterized in that the parking planning analysis is performed on the real-time parking groups and parking demand trends to obtain the initial parking space allocation information, including: By collecting data such as the entry and exit time, parking location, and parking duration of vehicles in the parking lot, and combining it with the license plate number, a vehicle feature database is established. Based on the data content of the vehicle feature database, the system uses a time series mining algorithm to extract the periodic travel patterns of vehicles, including the parking time distribution on weekdays and holidays, vehicle parking location preferences, and other regularity data.

[0048] Based on the data of vehicle travel patterns, the entropy value of parking demand trend is calculated. Specifically, the parking lot is divided into multiple regional units, the parking demand of each regional unit in different periods is counted, and the entropy value of parking demand in each area is calculated using information entropy theory. The entropy value reflects the uncertainty of regional parking demand. The larger the entropy value, the greater the volatility of parking demand.

[0049] For the obtained parking entropy distribution data, a two-way propagation neural network model is used for dynamic prediction. The model calculates the parking demand of each area in the future period through forward propagation, continuously optimizes the prediction parameters through back propagation, and finally outputs the parking space occupancy prediction data of each area in different time periods.

[0050] Based on the dynamic prediction of parking spaces, a multi-level resource allocation model is established according to factors such as the geographical location, parking space type, and usage frequency of different regions. The model divides parking space resources into different levels such as core areas, sub-core areas, and general areas, and assigns corresponding configuration weights to each level. The weight setting comprehensively considers multiple dimensions such as regional importance and parking space utilization efficiency.

[0051] The improved particle swarm algorithm is used to optimize the hierarchical configuration weight data. By setting multiple constraints such as minimizing parking time, minimizing walking distance, and maximizing resource utilization, the algorithm continuously adjusts the configuration weights of each area during the iterative optimization process, and finally obtains the optimal parking configuration solution that meets multiple constraints.

[0052] After completing the parking space optimization configuration, a global dynamic balance model is established. The model monitors the parking space usage in each area in real time, calculates the load balance between areas, and dynamically adjusts the parking space allocation plan based on the calculation results. When a certain area is saturated with parking spaces or the vacancy rate is too high, the parking space allocation strategy of the adjacent area will be automatically adjusted to ensure the balanced utilization of the entire parking lot resources. After the global dynamic balance calculation, the system generates an initial parking space allocation plan containing information such as specific parking space numbers, expected parking periods, and recommended routes.

[0053] This embodiment establishes a vehicle feature database and extracts travel rules in combination with a time series mining algorithm. The parking guidance method achieves accurate grasp of vehicle parking behavior and effectively improves the accuracy of parking demand prediction. The information entropy theory is used to quantify the parking demand, so that the system can objectively evaluate the fluctuation characteristics of parking demand in each region, providing reliable data support for subsequent resource allocation. The two-way propagation neural network is used for dynamic prediction, which significantly improves the adaptability and prediction accuracy of the prediction model to emergencies. The parking spaces are hierarchically managed through a multi-level resource allocation model, which realizes the differentiated configuration of parking resources and improves the service level of key areas. The improved particle swarm algorithm is combined for multi-constraint optimization, which ensures the scientificity and rationality of the parking space allocation plan while meeting multiple goals such as minimizing parking time and minimizing walking distance. A global dynamic balance model is established to monitor and adjust the parking space usage in real time, effectively avoiding the problem of unbalanced resource allocation between regions and improving the overall operation efficiency and service quality of the parking lot.

[0054] In one embodiment, real-time parking demand information of a parking lot is obtained, parking prediction is performed based on the parking demand trend, and a corresponding parking guidance plan is obtained, including: The real-time parking demand information of the parking lot is collected and obtained through multiple data collection terminals, which include license plate recognition systems, geomagnetic detectors, ultrasonic sensors and other equipment. These devices collect and upload data such as vehicle entry and exit information and parking space occupancy status information in real time.

[0055] When performing periodic decomposition on the collected parking demand information, the wavelet transform method is used to decompose the time series data into three parts: trend item, period item and random item. The trend item reflects the long-term trend of parking demand, the period item reflects the periodic laws such as daily change and weekly change, and the random item represents irregular fluctuation. The segmented feature sequence obtained by decomposition contains key indicators such as time label, vehicle flow, and occupancy rate.

[0056] When calculating the parking correlation index based on the segmented feature sequence, the Pearson correlation coefficient is used to evaluate the correlation degree of parking demand between different time periods and different areas. The correlation index includes two dimensions: time correlation (reflecting the correlation degree of adjacent time periods) and spatial correlation (reflecting the correlation degree of adjacent areas).

[0057] In the deep fusion process of parking correlation indicators, a deep neural network model is used to extract and combine the features of time correlation and spatial correlation. The input layer of the neural network receives the correlation indicator data, and finally outputs the regional parking weight matrix through nonlinear transformation of multiple hidden layers. Each element in the matrix represents the parking demand weight of the corresponding area in a specific period of time.

[0058] The calculation of regional parking density distribution is based on the kernel density estimation method, which transforms the weight matrix into a continuous probability density function. This function describes the spatial distribution of parking demand, and a higher density value indicates a greater parking demand in the area.

[0059] The K-means clustering algorithm is used to extract the regional parking regularity characteristics, and the areas with similar parking density distribution characteristics are classified into one category. The number of clusters is determined according to the actual area division of the parking lot, and the clustering results reflect the parking behavior patterns in different areas.

[0060] The parking flow prediction uses the long short-term memory network (LSTM) model, which can effectively capture the long-term dependencies in time series data. The prediction model uses historical parking data as training samples, comprehensively considers time characteristics, weather conditions, surrounding activities and other factors, and outputs the parking flow prediction value for the future period.

[0061] In the stage of intelligent parking resource configuration, the optimal parking guidance plan is formulated based on the predicted parking flow data, combined with the actual parking lot capacity and road conditions. The plan includes specific measures such as diversion strategies for each entrance, dynamic guidance information release, and parking space pre-allocation to achieve balanced utilization of parking resources.

[0062] The execution of the guidance plan is realized through the parking lot management system. The system pushes the guidance information to various display screens, mobile terminals and other devices to provide car owners with real-time parking location recommendations and route navigation services. The system continuously monitors the execution effect of the plan, adjusts and optimizes the guidance strategy in a timely manner, and ensures the efficient use of parking resources.

[0063] This embodiment uses a variety of data collection terminals to obtain parking lot information in real time, and combines the wavelet transform method to periodically decompose parking demand, which can fully and accurately grasp the dynamics of parking lots and improve the reliability and accuracy of data collection. By introducing the Pearson correlation coefficient to evaluate the parking correlation of different dimensions, and using deep neural networks for feature fusion, the parking demand analysis is more scientific and comprehensive, and the accuracy of decision-making is improved. The kernel density estimation method is used to calculate the parking density distribution, and the K-means clustering algorithm is combined to extract regional features, which can effectively identify parking hot spots and optimize resource allocation efficiency. The LSTM model is used to predict parking flow, and a variety of influencing factors are comprehensively considered to improve the accuracy and reliability of the prediction. Based on the prediction results, an intelligent parking guidance plan is formulated, and the guidance information is pushed in real time through the management system, which not only improves the efficiency of parking lot use, but also improves the user's parking experience, and realizes the intelligent and precise adjustment of parking resources.

[0064] In one embodiment, the initial parking space allocation information and the parking guidance plan are subjected to parking space guidance analysis to obtain global parking guidance information, including: In the density distribution calculation process, the initial parking space allocation information is subjected to spatial density calculation using the kernel density estimation algorithm to generate a heat map reflecting the spatial distribution characteristics of parking spaces. The heat map smoothes the parking space location data based on the Gaussian kernel function to form a continuous density distribution surface, where the color depth indicates the density of parking spaces, providing a reference for spatial distribution for subsequent optimization. The bandwidth parameter of the kernel density estimation is adaptively adjusted according to the spatial distribution characteristics of the parking space data to ensure that the heat map accurately reflects the parking space distribution status.

[0065] In the multi-objective optimization stage, the parking space allocation balance, driving distance minimization, and traffic flow dispersion are taken as optimization objectives, and the Pareto optimal criterion is used to construct a multi-objective optimization model. The parking space spatial distribution heat map is introduced as a constraint condition in the optimization process, and the optimal guidance strategy is solved by genetic algorithm. This strategy ensures the rational allocation and utilization of parking space resources while satisfying multiple objective functions. The weight coefficients of each objective function in the optimization model are determined by the hierarchical analysis method to achieve a dynamic balance between the objectives.

[0066] In the multi-source shortest path calculation stage, the guidance strategy is planned based on the Dijkstra algorithm, and a dynamic obstacle avoidance mechanism is built in combination with real-time road condition information. This mechanism dynamically updates the road network weights and generates the optimal obstacle avoidance guidance path through real-time monitoring of factors such as road congestion and temporary obstacles. A hierarchical processing strategy is adopted in the path planning process to establish a coordination mechanism between the local path and the global path to improve planning efficiency.

[0067] In the traffic flow entropy analysis, the entropy of the traffic flow on the dynamic obstacle avoidance guidance path is calculated to evaluate the load of the path. The entropy calculation is based on the information entropy theory. By measuring the uncertainty of the spatiotemporal distribution of traffic flow, a load index reflecting the congestion of the path is obtained. This index fully considers the time-varying characteristics and spatial distribution characteristics of traffic flow, providing a quantitative basis for path optimization.

[0068] In the joint iterative optimization process, the path load index is coupled with the initial parking allocation information for analysis, and the allocation scheme is optimized through alternating iterations. The iterative process adopts an adaptive step control strategy to dynamically adjust parameters according to the convergence of the optimization target to ensure the stability of the algorithm. The optimization results form an adaptive scheduling scheme to achieve the coordinated optimization of parking allocation and path planning.

[0069] In the regional balance factor mapping stage, the parking pressure index of each area is calculated based on the adaptive scheduling scheme, and the regional balance factor mapping relationship is established. The mapping process takes into account factors such as regional parking capacity, surrounding road traffic capacity, and historical parking demand to generate global parking guidance information. This information represents the guidance priority between regions in the form of a matrix, providing support for parking management decisions.

[0070] This embodiment uses a kernel density estimation algorithm to calculate the spatial density of the initial parking space allocation information, generates a parking space spatial distribution heat map, and realizes the accurate quantitative expression of the parking resource distribution characteristics, providing reliable data support for subsequent optimization decisions. The guidance strategy formulation process based on the multi-objective optimization model comprehensively considers multiple objectives such as parking space balance, driving distance and traffic flow, while ensuring the reasonable allocation of parking resources, effectively reducing the overall operating cost of the system. By combining the dynamic obstacle avoidance mechanism with the multi-source shortest path algorithm, real-time response to road condition changes is achieved, and the adaptability and reliability of the guidance scheme are improved. The path load is evaluated by using the traffic flow entropy analysis method, and combined with the joint iterative optimization technology, an adaptive scheduling scheme is constructed, which realizes the coordinated optimization of parking space allocation and path planning, and significantly improves the efficiency of parking management. The global parking guidance information generation method based on regional balance factor mapping provides a quantitative basis for parking management decisions and effectively improves the utilization rate of parking resources.

[0071] Reference Figure 2 As shown, the present invention further provides a parking guidance system based on data analysis, which is applied to any one of the above-mentioned parking guidance methods based on data analysis, comprising: The acquisition module is used to obtain the parking space data of the parking lot for regional division and obtain the corresponding regional parking space information; An analysis module is used to obtain the user's real-time parking demand information, associate the parking demand information with the regional parking space information, and obtain the corresponding real-time parking group; The association module is used to obtain the historical parking data of the parking lot to predict the parking demand and obtain the corresponding parking demand trend; A processing module, which is used to perform parking planning analysis on real-time parking groups and parking demand trends to obtain initial parking space allocation information; A control module is used to obtain real-time parking demand information of the parking lot, make parking predictions based on the parking demand trend, and obtain corresponding parking guidance plans; The execution module is used to perform parking guidance analysis on the initial parking space allocation information and the parking guidance plan to obtain global parking guidance information.

[0072] The present invention provides a parking guidance system based on data analysis. By dividing parking lot data into regions and analyzing parking spaces, combined with real-time user demand information, the parking resource distribution status in different regions can be more accurately evaluated, thereby improving the accuracy of parking guidance and providing a more scientific basis for parking space allocation. By correlating and analyzing user demand data with regional parking space information, refined management of parking resources in different regions can be achieved, which is helpful to achieve reasonable allocation and avoid waste of resources. Trend analysis of parking demand based on historical data can ensure that the system can operate efficiently in different time periods and scenarios, reduce vacant parking spaces, and improve the overall utilization rate of parking lots. By comprehensively analyzing real-time parking groups and demand trends, a more reasonable parking guidance plan is formulated, and the optimization operation of the entire parking lot is achieved through a global adjustment strategy, thereby effectively utilizing parking resources and reducing congestion and queuing. And by considering the impact of real-time demand on parking guidance, the parking space allocation strategy can be flexibly adjusted according to the characteristics of different time periods and demand changes, so that the system is more adaptable to dynamically changing parking scenarios.

[0073] It should be noted that technicians in the relevant technical field can clearly understand that for the convenience and conciseness of description, the specific working process of the system and each module described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0074] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A parking guidance method based on data analysis, characterized in that: include: Obtain parking space data of the parking lot for regional division and obtain corresponding regional parking space information; Acquire the user's real-time parking demand information, associate the parking demand information with the regional parking space information, and obtain a corresponding real-time parking group; Acquire historical parking data of the parking lot to perform parking demand prediction and obtain corresponding parking demand trends; Performing parking planning analysis on the real-time parking group and the parking demand trend to obtain initial parking space allocation information; Acquire the real-time parking demand information of the parking lot, perform parking prediction based on the parking demand trend, and obtain a corresponding parking guidance plan; The initial parking space allocation information and the parking guidance plan are subjected to parking space guidance analysis to obtain global parking guidance information.

2. The parking guidance method based on data analysis according to claim 1, characterized in that: The method of obtaining parking space data of the parking lot and dividing the parking space into regions to obtain corresponding regional parking space information includes: Performing spatial grid division on the parking space data to obtain initial grid unit data; Performing multi-layer grid index construction on the initial grid unit data to obtain a hierarchical grid index structure; Calculating the grid parking space density on the hierarchical grid index to obtain grid unit parking space density distribution data; Clustering adjacent grids according to the grid unit parking space density distribution data to obtain an initial area division result; Perform parking space coding mapping on the initial area division result to obtain parking space coding information in the area; Data matching is performed according to the parking space coding information in the area to obtain the parking space information in the area.

3. The parking guidance method based on data analysis according to claim 1, characterized in that: The acquiring of the user's real-time parking demand information, associating the parking demand information with the regional parking space information, and obtaining a corresponding real-time parking group includes: Extract information from parking requests submitted by users to obtain real-time parking demand information; Performing area matching screening according to the real-time parking demand information to obtain a set of candidate parking areas; Performing parking space availability analysis on the candidate parking area set to obtain real-time vacant parking space data; Calculate the distance of each candidate area according to the real-time vacant parking space data to obtain the shortest path data; Prioritizing the shortest path data and the real-time vacant parking space data to obtain an initial parking space allocation sequence; Performing time conflict detection on the initial parking space allocation sequence to obtain a valid parking space combination; Demand grouping is performed according to the effective parking space combination to obtain the real-time parking group.

4. The parking guidance method based on data analysis according to claim 1, characterized in that: The acquiring of the historical parking data of the parking lot to perform parking demand prediction and obtain a corresponding parking demand trend includes: Performing sequence decomposition on the historical parking data to obtain a parking feature sequence; Perform parking periodicity analysis according to the parking feature sequence to obtain time distribution characteristics; Performing clustering processing on the time distribution characteristics to obtain parking regularity characteristics; Performing parking correlation analysis according to the parking regularity characteristics to obtain location correlation characteristics; Performing demand forecasting on the location-related features to obtain parking demand change rules; The parking demand trend is obtained by performing trend analysis on the parking feature sequence according to the parking demand change rule.

5. The parking guidance method based on data analysis according to claim 1, characterized in that: The performing parking planning analysis on the real-time parking group and the parking demand trend to obtain initial parking space allocation information includes: Extracting vehicle information features from the real-time parking group to obtain vehicle travel pattern data; Performing entropy calculation processing on the parking demand trend according to the vehicle travel pattern data to obtain parking entropy distribution data; Performing bidirectional dynamic propagation prediction on the parking entropy value distribution data to obtain parking space dynamic prediction data; Performing hierarchical allocation of parking space resources according to the dynamic prediction data of parking spaces to obtain hierarchical configuration weight data; Performing multi-constraint group intelligent optimization on the hierarchical configuration weight data to obtain parking space optimization configuration data; A global dynamic balance calculation is performed on the parking space optimization configuration data to obtain the initial parking space allocation information.

6. The parking guidance method based on data analysis according to claim 1, characterized in that: The acquiring of the real-time parking demand information of the parking lot, performing parking prediction based on the parking demand trend, and obtaining a corresponding parking guidance plan includes: Performing periodic decomposition on the parking demand information to obtain a segmented feature sequence; Calculating the parking demand trend index according to the segmented feature sequence to obtain a parking correlation index; Deeply integrating the parking correlation indicators to obtain a regional parking weight matrix; Calculate the parking demand probability distribution according to the regional parking weight matrix to obtain the regional parking density distribution; Performing group clustering on the parking density distribution of the area to obtain regional parking regularity characteristics; Predicting parking flow according to the parking pattern characteristics of the area to obtain dynamic parking prediction data; Intelligently configure parking resources based on the dynamic parking prediction data to obtain a parking guidance plan.

7. The parking guidance method based on data analysis according to claim 1, characterized in that: The performing parking guidance analysis on the initial parking space allocation information and the parking guidance scheme to obtain global parking guidance information includes: Performing density distribution calculation on the initial parking space allocation information to obtain a parking space space distribution heat map; Perform multi-objective optimization on the parking guidance scheme according to the parking space space distribution heat map to obtain a guidance strategy; Performing multi-source shortest path calculation on the guidance strategy to obtain a dynamic obstacle avoidance guidance path; Performing traffic flow entropy analysis based on the dynamic obstacle avoidance guidance path to obtain a path load index; Performing joint iterative optimization on the path load index and the initial parking space allocation information to obtain an adaptive scheduling solution; Regional balance factor mapping is performed according to the adaptive scheduling scheme to obtain global parking guidance information.

8. A parking guidance system based on data analysis, characterized in that: The parking guidance method based on data analysis applied to any one of claims 1 to 7 above comprises: A collection module, which is used to obtain parking space data of the parking lot, divide the area into regions, and obtain corresponding regional parking space information; An analysis module, the analysis module is used to obtain the user's real-time parking demand information, associate the parking demand information with the regional parking space information, and obtain a corresponding real-time parking group; A correlation module, the correlation module is used to obtain historical parking data of the parking lot to perform parking demand prediction and obtain a corresponding parking demand trend; A processing module, the processing module is used to perform parking planning analysis on the real-time parking group and the parking demand trend to obtain initial parking space allocation information; A control module, the control module is used to obtain real-time parking demand information of the parking lot, perform parking prediction based on the parking demand trend, and obtain a corresponding parking guidance plan; An execution module is used to perform parking guidance analysis on the initial parking space allocation information and the parking guidance plan to obtain global parking guidance information.

Images