Intelligent parking guidance method, system and storage medium

Abstract

The present application relates to the field of data processing technology, and discloses an intelligent parking guidance method, system, and storage medium. The method includes: performing time series analysis and pattern recognition processing on the target traffic data set to obtain traffic flow forecast data and parking demand forecast data; performing multi-objective optimization processing on the parking demand forecast data to obtain target decision data and converting and prioritizing the data to obtain real-time information data suitable for different publishing platforms; performing correlation analysis and path planning processing on user parking behavior data, real-time traffic status data, and parking lot occupancy rate data to obtain candidate parking recommendation path data; performing multi-dimensional cross-analysis processing on user feedback data and occupancy rate data to obtain evaluation index data, and performing data correction on the candidate parking recommendation path data to obtain target parking recommendation path data. The present application improves the timeliness and accuracy of intelligent parking guidance.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to an intelligent parking guidance method, system and storage medium. Background Art

[0002] With the acceleration of urbanization and the rapid growth of private car ownership, parking difficulties have become increasingly prominent, becoming a major challenge plaguing modern urban development. To address this issue, various parking guidance systems have emerged. Existing parking guidance systems typically use fixed information dissemination devices, such as electronic display screens or mobile applications, to provide drivers with information on the location and availability of nearby parking lots. These systems collect real-time parking occupancy data and transmit this information to a central control center for processing and distribution, helping drivers quickly find available parking spaces.

[0003] However, existing parking guidance systems have numerous shortcomings. First, these systems often only provide static or semi-static parking information, failing to fully account for real-time traffic conditions, user preferences, and the dynamic changes in parking needs. Second, existing systems lack in-depth analysis of user historical behavior and feedback, making it difficult to provide personalized parking recommendations. Furthermore, these systems typically treat parking lots as isolated entities, ignoring the interplay between parking behavior and surrounding factors such as traffic flow and commercial activities. Finally, existing systems are inefficient in data processing and information dissemination, making it difficult to cope with the demands of rapidly processing and accurately delivering large-scale real-time data. Summary of the Invention

[0004] The present application provides an intelligent parking guidance method, system and storage medium for improving the timeliness and accuracy of intelligent parking guidance.

[0005] In the first aspect, the present application provides an intelligent parking guidance method, which includes: real-time collection and standardized processing of multi-source traffic data to obtain a target traffic data set containing traffic flow, parking space availability, environmental factors and historical trends; time series analysis and pattern recognition processing of the target traffic data set to obtain traffic flow forecast data and parking demand forecast data; based on the traffic flow forecast data, multi-objective optimization processing of the parking demand forecast data to obtain target decision data, wherein the target decision data includes: traffic flow allocation plan, parking resource allocation strategy and vehicle path recommendation data; multi-channel information format conversion and priority sorting processing of the target decision data to obtain real-time information data suitable for different publishing platforms; real-time collection of user parking Behavior data, real-time traffic status data and updated parking lot occupancy data, and perform correlation analysis and path planning processing on the user parking behavior data, the real-time traffic status data and the parking lot occupancy data to obtain candidate parking recommendation path data including the target parking lot and the navigation route; perform multi-dimensional cross-analysis processing on the real-time collected user feedback data and the occupancy data to obtain evaluation index data, wherein the evaluation index data include: user satisfaction and resource utilization efficiency; perform data correction on the candidate parking recommendation path data through the evaluation index data to obtain target parking recommendation path data, wherein the target parking recommendation path data include: target parking lot geographic coordinates, target driving route coordinate sequence, estimated driving time, estimated parking fee, and real-time traffic data.

[0006] In combination with the first aspect, in a first implementation of the first aspect of the present application, the multi-source traffic data includes road vehicle density data, vehicle passage data at parking lot entrances and exits, weather condition data, special event information, and historical parking data. The real-time collection and standardized processing of the multi-source traffic data to obtain a target traffic data set containing traffic flow, parking space availability, environmental factors, and historical trends includes: performing dynamic threshold segmentation processing on the road vehicle density data to obtain traffic flow classification data reflecting traffic flow; performing difference calculation processing on the vehicle passage data at the parking lot entrances and exits to obtain real-time parking space occupancy data; performing semantic analysis processing on the weather condition data and the special event information to obtain environmental impact factor data; performing periodic decomposition processing on the historical parking data to obtain parking trend cycle data; performing spatiotemporal correlation analysis on the traffic flow classification data and the real-time parking space occupancy rate data to obtain parking demand hotspot distribution data; performing weighted fusion processing on the environmental impact factor data and the parking trend cycle data to obtain basic parking behavior prediction data; performing density clustering processing on the parking demand hotspot distribution data to obtain regional parking pressure assessment data; performing time series interpolation processing on the parking behavior prediction basic data to obtain parking demand prediction data in the continuous time domain; performing adaptive weight allocation processing on the regional parking pressure assessment data and the parking demand prediction data to obtain comprehensive parking index data; performing multi-scale normalization processing on the comprehensive parking index data to obtain a target traffic data set containing traffic flow, parking space availability, environmental factors and historical trends.

[0007] In combination with the first aspect, in a second implementation of the first aspect of the present application, the target traffic data set is subjected to time series analysis and pattern recognition processing to obtain traffic flow prediction data and parking demand prediction data, including: performing time window segmentation processing on the target traffic data set to obtain parking data segments of multiple time granularities, and performing Fourier transform processing on the parking data segments of multiple time granularities to obtain parking periodic feature data; performing wavelet decomposition processing on the parking periodic feature data to obtain multi-scale parking trend data, and performing time series prediction processing on the multi-scale parking trend data through a long short-term memory network algorithm to obtain initial parking demand prediction data; the initial parking demand prediction data and real-time parking demand prediction data are processed. The collected real-time traffic flow data is subjected to correlation analysis to obtain a flow-demand correlation factor, and the flow-demand correlation factor is subjected to dynamic weight allocation to obtain a parking demand adjustment parameter; the initial parking demand forecast data and the parking demand adjustment parameter are adaptively fused to obtain optimized parking demand forecast data, and trend extraction and seasonal decomposition are performed on historical vehicle flow data to obtain vehicle flow basic characteristic data; anomaly detection is performed on the vehicle flow basic characteristic data and real-time traffic event data to obtain a vehicle flow fluctuation factor, and the vehicle flow basic characteristic data and the vehicle flow fluctuation factor are combined for forecasting to obtain the vehicle flow forecast data and the parking demand forecast data.

[0008] In combination with the first aspect, in a third implementation of the first aspect of the present application, based on the vehicle flow prediction data, multi-objective optimization processing is performed on the parking demand prediction data to obtain target decision data, wherein the target decision data includes: a traffic flow allocation plan, a parking resource allocation strategy, and vehicle path recommendation data, including: performing network flow model construction processing on the vehicle flow prediction data to obtain a traffic flow network topology structure, and performing spatial clustering processing on the parking demand prediction data to obtain parking demand hotspot area data; performing matching calculation processing on the traffic flow network topology structure and the parking demand hotspot area data to obtain an initial traffic flow allocation plan, and performing capacity constraint optimization processing on the initial traffic flow allocation plan to obtain the traffic flow allocation plan; Performing difference analysis on existing parking resource data and the parking demand forecast data to obtain parking resource gap data, and dynamically allocating the parking resource gap data to obtain an initial parking resource allocation strategy; performing conflict detection on the initial parking resource allocation strategy to obtain the parking resource allocation strategy; fusing the real-time road network status data and the traffic flow allocation plan to obtain a multi-dimensional path evaluation index; performing path search on the multi-dimensional path evaluation index using a particle swarm optimization algorithm to obtain a candidate vehicle path set; performing multi-objective trade-off processing on the candidate vehicle path set to obtain the vehicle path recommendation data, and integrating the traffic flow allocation plan, the parking resource allocation strategy, and the vehicle path recommendation data into the target decision data.

[0009] In combination with the first aspect, in a fourth implementation of the first aspect of the present application, the target decision data is subjected to multi-channel information format conversion and priority sorting processing to obtain real-time information data suitable for different publishing platforms, including: data compression processing on the traffic flow allocation plan to obtain traffic flow summary data; spatial index construction processing on the parking resource allocation strategy to obtain parking space distribution data; path simplification processing on the vehicle path recommendation data to obtain key path node data; data fusion processing on the traffic flow summary data, the parking space distribution data and the key path node data to obtain an information data packet; multi-level cache strategy processing on the information data packet to obtain a hierarchically stored information cache structure, and real-time update frequency analysis processing on the information cache structure to obtain information timeliness scoring data; priority sorting processing on the information timeliness scoring data to obtain an information release queue, and multi-channel adaptation processing on the data in the information release queue to obtain formatted information for different publishing platforms; and secure encryption processing on the formatted information for different publishing platforms to obtain the real-time information data suitable for different publishing platforms.

[0010] In combination with the first aspect, in the fifth implementation method of the first aspect of the present application, the real-time collection of user parking behavior data, real-time traffic condition data and parking lot occupancy data, and the correlation analysis and path planning processing of the user parking behavior data, the real-time traffic condition data and the parking lot occupancy data are performed to obtain candidate parking recommendation path data including the target parking lot and the navigation route, including: clustering analysis processing of the user parking behavior data to obtain user parking preference information, and road congestion calculation processing of the traffic condition data to obtain a traffic smoothness index; difference statistical processing of the parking lot occupancy data to obtain a dynamic parking space occupancy; correlation analysis of the user parking preference signal and the dynamic parking space occupancy The method comprises the steps of: performing a correlation analysis process on the traffic flow index to obtain a personalized parking lot recommendation list; performing a threshold segmentation process on the traffic flow index to obtain a set of feasible driving sections; performing a path planning process on the set of feasible driving sections using a heuristic search algorithm to obtain a plurality of candidate navigation routes; performing a geographic coordinate extraction process on the parking lots in the personalized parking lot recommendation list to obtain a target parking lot coordinate set; performing a matching degree calculation process on the plurality of candidate navigation routes and the target parking lot coordinate set to obtain a route-parking lot matching solution; performing a comprehensive scoring process on the route-parking lot matching solution to obtain an optimal parking recommendation path; and performing data encapsulation process on the optimal parking recommendation path to obtain candidate parking recommendation path data including a target parking lot and a navigation route.

[0011] In combination with the first aspect, in the sixth implementation method of the first aspect of the present application, the user feedback data and the occupancy rate data collected in real time are subjected to multi-dimensional cross-analysis processing to obtain evaluation index data, wherein the evaluation index data include: user satisfaction and resource utilization efficiency, including: performing sentiment analysis processing on the user feedback data to obtain an initial parking experience satisfaction score, and performing time decay processing on the initial parking experience satisfaction score to obtain timeliness weighted satisfaction data; performing time series decomposition processing on the parking lot occupancy rate data to obtain parking lot utilization efficiency trend data, and performing correlation analysis processing on the timeliness weighted satisfaction data and the parking lot utilization efficiency trend data to obtain satisfaction-efficiency correlation index; extract distribution features of user parking time data to obtain parking time pattern data, and calculate the matching degree of the parking time pattern data and parking lot turnover rate data to obtain a time-space utilization efficiency index; fuse the satisfaction-efficiency correlation index and the time-space utilization efficiency index to obtain a comprehensive evaluation initial value, and perform statistical analysis on the parking guidance response time data to obtain a response efficiency score; perform weighted average processing on the comprehensive evaluation initial value and the response efficiency score to obtain the original data of the target evaluation index; normalize and quantify the original data of the target evaluation index to obtain the evaluation index data, wherein the evaluation index data includes: user satisfaction and resource utilization efficiency.

[0012] In combination with the first aspect, in the seventh implementation method of the first aspect of the present application, the candidate parking suggestion path data is corrected by the evaluation index data to obtain target parking suggestion path data, wherein the target parking suggestion path data includes: the geographical coordinates of the target parking lot, the target driving route coordinate sequence, the expected driving time, the expected parking fee, and the real-time road condition data, including: performing threshold analysis processing on the user satisfaction in the evaluation index data to obtain a parking lot satisfaction ranking list; performing geographic information extraction processing on the target parking lot in the candidate parking suggestion path data to obtain a candidate parking lot coordinate set, and performing matching screening processing on the parking lot satisfaction ranking list and the candidate parking lot coordinate set to obtain the target parking lot geographical coordinates; performing path analysis on the navigation route in the candidate parking suggestion path data Smoothing is performed to obtain an initial driving route coordinate sequence, and the initial driving route coordinate sequence is optimized to obtain a target driving route coordinate sequence; the target driving route coordinate sequence is segmented to obtain a road section unit set, and the road section unit set and real-time traffic flow data are fused and analyzed to obtain the estimated travel time of each road section; the estimated travel time of each road section is accumulated and calculated to obtain the estimated travel time; the historical price data of the target parking lot is time series predicted to obtain the estimated parking fee; the collected road condition data is semantically processed to obtain the real-time road condition data, and the geographical coordinates of the target parking lot, the target driving route coordinate sequence, the estimated travel time, the estimated parking fee and the real-time road condition data are integrated into the target parking recommended path data.

[0013] In a second aspect, the present application provides an intelligent parking guidance system, the intelligent parking guidance system comprising:

[0014] A processing module is used to collect and standardize multi-source traffic data in real time to obtain a target traffic dataset containing traffic flow, parking space availability, environmental factors and historical trends;

[0015] an identification module for performing time series analysis and pattern recognition processing on the target traffic data set to obtain traffic flow data and parking demand prediction data;

[0016] an optimization module, configured to perform multi-objective optimization processing on the parking demand prediction data based on the vehicle flow data to obtain target decision data, wherein the target decision data includes: a traffic flow allocation plan, a parking resource allocation strategy, and vehicle path recommendation data;

[0017] A conversion module is used to convert the target decision data into multiple information formats and perform priority sorting processing to obtain real-time information data suitable for different publishing platforms;

[0018] a planning module for collecting real-time user parking behavior data, real-time traffic status data, and parking lot occupancy rate data, and performing correlation analysis and path planning on the user parking behavior data, the real-time traffic status data, and the parking lot occupancy rate data to obtain candidate parking path data including a target parking lot and a navigation route;

[0019] An analysis module is used to perform multi-dimensional cross-analysis processing on the user feedback data and the occupancy rate data collected in real time to obtain evaluation index data, wherein the evaluation index data includes: user satisfaction and resource utilization efficiency;

[0020] A correction module is used to correct the candidate parking suggestion path data using the evaluation index data to obtain target parking suggestion path data, wherein the target parking suggestion path data includes: the geographical coordinates of the target parking lot, the coordinate sequence of the target driving route, the estimated driving time, the estimated parking fee, and real-time traffic data.

[0021] A third aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned intelligent parking guidance method.

[0022] The technical solution provided in this application utilizes real-time collection and standardized processing of multi-source traffic data to generate a target traffic dataset encompassing traffic flow, parking availability, environmental factors, and historical trends. This not only improves data timeliness but also enhances decision-making reliability. Time series analysis and pattern recognition are performed on the target traffic dataset to generate traffic flow data and parking demand forecast data, enabling prediction of future traffic conditions and parking demand, thereby enabling preemptive and appropriate resource allocation. Multi-objective optimization is performed on parking demand forecast data based on traffic flow data to generate target decision data, including traffic flow allocation plans, parking resource allocation strategies, and vehicle route recommendations. This target decision data is then converted into multi-channel information formats and prioritized to generate real-time information data suitable for different publishing platforms. This significantly improves the efficiency and reach of information dissemination, enabling users to access parking guidance information across various devices and platforms. Real-time collection of user behavior data, real-time traffic conditions data, and parking lot occupancy data, coupled with correlation analysis and route planning, generates candidate parking recommendation routes, including target parking lots and navigation routes. This achieves precise response to user needs while also taking into account real-time road conditions, improving the accuracy and practicality of recommendations. Multi-dimensional cross-analysis and processing are performed on the user feedback data and occupancy data collected in real time to obtain evaluation index data including user satisfaction and resource utilization efficiency. The candidate parking recommendation path data is corrected based on the evaluation index data to obtain the target parking lot geographical coordinates, target driving route coordinate sequence, estimated driving time, estimated parking fee and real-time road condition data. Through multi-step data processing and decision optimization, the entire process from data collection to end-user recommendation is intelligentized, which not only greatly improves parking efficiency, reduces the time and cost of users looking for parking spaces, but also improves the timeliness and accuracy of intelligent parking guidance. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0024] Figure 1 This is a schematic diagram of an embodiment of the intelligent parking guidance method in the embodiment of the present application;

[0025] Figure 2 This is a schematic diagram of an embodiment of the intelligent parking guidance system in the embodiment of the present application. DETAILED DESCRIPTION

[0026] The embodiments of the present application provide an intelligent parking guidance method, system and storage medium. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0027] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 , an embodiment of the intelligent parking guidance method in the embodiment of the present application includes:

[0028] Step S101: Real-time collection and standardization of multi-source traffic data to obtain a target traffic dataset containing traffic flow, parking space availability, environmental factors, and historical trends;

[0029] It is understandable that the execution subject of this application can be the intelligent parking guidance system, or it can be a terminal or a server, and the specific implementation is not limited here. The embodiment of this application is described by taking the server as the execution subject as an example.

[0030] Specifically, real-time traffic data is collected from multiple sources, including road sensors, cameras, vehicle GPS devices, parking management systems, weather stations, and historical databases. The first step in standardization is data cleaning to remove outliers and noise. For example, traffic flow data may contain abnormal readings due to equipment failures, which need to be identified and eliminated through statistical methods. Next, data from different sources is synchronized temporally and spatially aligned to ensure that all data is analyzed within the same spatiotemporal framework. When processing traffic flow data, vehicle counts and speed measurements are used to quantify traffic flow conditions on the road. This raw data is aggregated and averaged to convert hourly lane-by-lane traffic volume and average speed. Parking space availability data is calculated by real-time monitoring of vehicle counts at parking lot entrances and exits, combined with the total number of parking spaces. Environmental factor data, such as weather conditions and air quality index, needs to be digitized and categorized for subsequent analysis. Time series analysis of data over a certain period of time identifies cyclical patterns and long-term trends. Techniques such as seasonal decomposition and trend extraction are employed to generate a dataset that reflects historical patterns.

[0031] All of this processed data is ultimately consolidated into a target traffic dataset. This dataset, in a standardized format, encompasses information on multiple dimensions, including traffic flow, parking availability, environmental factors, and historical trends. Each record in the dataset includes fields such as timestamp, location, and numerical values ​​for various indicators, facilitating subsequent analysis and decision-making.

[0032] For example, at a specific point in time, a record in the target traffic dataset may contain the following information: the real-time traffic volume on a main road is 1,200 vehicles per hour, with an average speed of 35 kilometers per hour; the available parking space rate of a nearby large parking lot is 15%; the current weather is light rain with visibility of 3 kilometers; according to historical data analysis, this point in time is usually the traffic peak period on weekdays, and the parking demand is 30% higher than usual.

[0033] Step S102: performing time series analysis and pattern recognition processing on the target traffic data set to obtain traffic flow forecast data and parking demand forecast data;

[0034] Specifically, performing time series analysis and pattern recognition on the target traffic dataset to generate traffic flow and parking demand forecasts is a key step in the intelligent parking guidance method. This process first involves data preprocessing, including data cleaning, denoising, and standardization. Time series analysis techniques are used to decompose the data and identify trends, seasonality, and cyclical patterns. Methods employed include moving average, exponential smoothing, and ARIMA (Autoregressive Integrated Moving Average) models.

[0035] For traffic flow forecasting, time series analysis can capture daily, weekly, and seasonal variations in traffic flow. For example, patterns such as weekday morning and evening rush hours and weekend shopping peaks can be identified through this analysis. Furthermore, considering the impact of special events (such as holidays and major events) on traffic flow, these factors can be incorporated as external variables. Parking demand forecasting combines time series analysis and pattern recognition techniques. In addition to considering the temporal trends of historical parking data, it is also necessary to analyze the correlation between parking behavior and other factors (such as weather and surrounding activities). Machine learning algorithms, such as random forests or support vector machines, may be used to identify key features that influence parking demand. Pattern recognition processing can identify recurring patterns and anomalies in complex traffic data. For example, clustering algorithms can be used to group similar traffic patterns and identify typical traffic condition types. These identified patterns are not only used for forecasting but also help understand the underlying laws of traffic behavior.

[0036] In practical applications, the latest real-time data is constantly used to update and calibrate the forecast to ensure its accuracy. At the same time, the forecast results are accompanied by a confidence interval to indicate the degree of uncertainty in the forecast.

[0037] For example, suppose we need to forecast parking demand for a commercial district. Hourly parking data from the past three months is collected, including information such as parking lot occupancy, surrounding road traffic volume, and weather conditions. Time series analysis identifies patterns on weekdays (e.g., a sharp increase in parking demand around 9:00 AM and 6:00 PM) and weekends (e.g., peak demand between 2:00 PM and 5:00 PM on Saturdays). Pattern recognition algorithms then correlate factors such as weather and surrounding events with parking demand, finding, for example, that parking demand increases by an average of 15% on rainy days. Finally, based on these analysis results, a forecast for parking demand is generated for the next 24 hours. For example, it predicts that parking demand will reach 85% of parking lot capacity at 3:00 PM tomorrow, Saturday, with a confidence interval of ±5%. This forecast incorporates the temporal patterns of historical data, the predicted weather conditions (sunny weather), and known information about surrounding events (mall promotions). This forecast provides accurate decision-making for the parking guidance system, helping to optimize parking resource allocation and reduce traffic congestion.

[0038] Step S103: Based on the traffic flow forecast data, multi-objective optimization processing is performed on the parking demand forecast data to obtain target decision data, wherein the target decision data includes: traffic flow allocation plan, parking resource allocation strategy and vehicle route recommendation data;

[0039] Specifically, the multi-objective optimization of parking demand forecast data based on traffic flow prediction is a complex and critical step in intelligent parking guidance. This process aims to balance traffic flow, parking demand, and resource utilization, ultimately generating target decision data that includes traffic flow allocation solutions, parking resource allocation strategies, and vehicle routing recommendations.

[0040] First, traffic flow forecast data and parking demand forecast data are integrated and analyzed. Traffic flow forecast data reflects the expected traffic conditions along different road segments in the road network, while parking demand forecast data indicates the expected usage of individual parking lots. Correlation analysis between these two sets of data can identify potential traffic bottlenecks and parking pressure points. Multi-objective optimization utilizes a variety of algorithms, including genetic algorithms, particle swarm optimization, and multi-criteria decision analysis. These algorithms simultaneously consider multiple objectives, such as minimizing overall traffic congestion, maximizing parking resource utilization, and minimizing the time vehicles spend searching for parking spaces. During the optimization process, each objective is assigned a different weight to reflect its relative importance. A dynamic traffic assignment model is used to generate traffic flow allocation plans. This model considers the predicted traffic flow and attempts to evenly distribute traffic across the network to avoid overcrowding on certain road segments. It also considers the distribution of parking demand, attempting to direct vehicles toward areas with sufficient parking spaces.

[0041] Developing a parking resource allocation strategy involves optimizing the allocation of existing parking resources. This includes dynamically adjusting parking rates to encourage drivers to use less-used parking lots or opening temporary parking areas during peak hours. Strategies also consider the specific needs of different types of vehicles, such as electric vehicles and vehicles for people with disabilities.

[0042] The generation of vehicle route recommendation data takes into account real-time traffic conditions, predicted traffic flow, and available parking space information. A shortest path algorithm (such as Dijkstra's algorithm) is used to calculate the optimal route, but "optimal" here takes into account not only distance but also expected travel time, congestion probability, and parking availability.

[0043] For example, suppose a commercial district is expected to see a large number of shoppers on Saturday afternoon. Traffic forecasts indicate that traffic on major roads will reach 150% of weekday levels, while parking demand forecasts indicate an 80% increase in parking demand. A multi-objective optimization process first generates a traffic flow allocation plan, directing some traffic to secondary roads to relieve pressure on the main arterials. It then develops parking resource allocation strategies, such as temporarily opening parking lots at nearby office buildings and offering preferential rates for parking lots farther from the commercial center. Finally, it generates personalized route recommendations for vehicles, taking into account predicted traffic conditions and parking space occupancy rates. For example, for a vehicle heading to the commercial district, the recommended route might avoid the expected congested main entrance, route via secondary roads to a more distant but ample parking lot, and provide walking directions from the parking lot to the destination.

[0044] Step S104: convert the target decision data into a multi-channel information format and perform priority sorting to obtain real-time information data suitable for different publishing platforms;

[0045] Specifically, data on traffic flow allocation plans, parking resource allocation strategies, and vehicle route recommendations are categorized and organized. Data is formatted based on the characteristics of different publishing platforms and user needs. This format conversion process considers multiple publishing platforms, such as mobile applications, in-car navigation systems, and roadside electronic displays. For mobile applications, data is converted into a lightweight JSON format for fast transmission and parsing. In-car navigation systems require a more streamlined data format, typically using binary encoding to reduce data volume. Roadside electronic displays convert information into concise text and graphic formats. Prioritization is a crucial step in ensuring that the most critical information reaches users promptly. The ranking algorithm considers factors such as the timeliness, importance, and relevance of the information. For example, warnings of severe traffic congestion or the sudden closure of a large parking lot are given higher priority. Furthermore, personalized recommendations are prioritized based on the user's location and destination.

[0046] The data processing process also involves information compression and optimization. For dynamic information that requires frequent updates, such as real-time parking space occupancy rates, an incremental update strategy is adopted, transmitting only the changed information, thereby reducing data transmission volume. For relatively stable information, such as parking lot locations, a caching mechanism is used to reduce repeated transmission. In practice, this process is dynamic and continuous. As new decision-making data is generated, the information conversion and sorting process continues, ensuring that users always have access to the latest and most relevant information. Simultaneously, the system monitors user responses and feedback, dynamically adjusting information release strategies to optimize the user experience.

[0047] For example, consider a busy commercial district where the target decision-making data includes traffic flow forecasts for the next two hours, expected occupancy rates for each parking lot, and recommended routes. The data processing process first categorizes this information: traffic flow forecasts are converted into a congestion level (1-5), parking lot occupancy rates are converted into percentages, and recommended routes are converted into a series of GPS coordinates. For mobile app users, the system generates a JSON-formatted data packet containing the congestion level within 500 meters of the user's current location, the occupancy forecasts for the five nearest parking lots, and an optimized route. This data packet is approximately 10KB and can be transmitted in less than one second on a 3G network. For roadside electronic signs, the system generates a concise message: "Road congestion 2 km ahead (Level 4). Recommend turning right to Parking Lot A (current vacancy 30%)." In the prioritization process, if the user is about to enter a severely congested area, the system immediately sends a warning message, taking precedence over regular parking updates.

[0048] Step S105: collecting user parking behavior data, real-time traffic status data, and updated parking lot occupancy rate data in real time, and performing correlation analysis and path planning on the user parking behavior data, real-time traffic status data, and parking lot occupancy rate data to obtain candidate parking path data including the target parking lot and navigation route;

[0049] It should be noted that user parking behavior data is collected through mobile applications and in-vehicle devices, including information such as users' historical parking locations, parking durations, and preferences. Real-time traffic condition data comes from road sensors, traffic cameras, and floating vehicle data, reflecting the current level of congestion and traffic speed on the road network. Parking lot occupancy data is obtained through each parking lot's real-time monitoring system, providing the current number of available parking spaces in each lot. After data collection, it undergoes data cleaning and preprocessing to remove outliers and noise. Correlation analysis is then performed on these three types of data. Correlation analysis uses data mining techniques, such as association rule mining algorithms, to identify potential relationships between user parking behavior and traffic conditions and parking lot occupancy rates. For example, it can analyze users' parking lot selection preferences under specific traffic conditions or user parking behavior patterns during periods of high occupancy.

[0050] Path planning is based on the results of association analysis and current real-time data. Using an improved A* algorithm or Dijkstra algorithm, multiple candidate paths are generated taking into account real-time traffic conditions, predicted travel time, and parking lot availability. Each path includes a recommended target parking lot and a navigation route to that parking lot. During the path generation process, the algorithm weighs multiple factors: travel time, the distance of the parking lot from the destination, the current and predicted availability of the parking lot, and the user's historical preferences. A comprehensive score is calculated for each candidate path through a multi-objective optimization method, and the paths with the highest scores are selected as candidate parking recommendation path data. As new data continues to flow in, the system will update the path recommendations in real time. For example, if it is detected that the original recommended route is suddenly congested, or the target parking lot is suddenly full, the system will quickly recalculate and provide a new recommended path.

[0051] For example: Suppose a user is driving to a shopping mall in the city center. The system first analyzes the user's historical parking data and finds that the user tends to choose a parking lot that is no more than 500 meters away from the destination on foot. Real-time traffic data shows that the current congestion index of the main road leading to the city center is 0.8 (0-1 scale, 1 is severe congestion). At the same time, parking lot occupancy data shows that the current occupancy rates of the three parking lots A, B, and C near the shopping mall are 85%, 70%, and 55%, respectively. Through association analysis, it is found that when the congestion index of the main road exceeds 0.7, the probability of the user choosing a route that is slightly longer but has smoother traffic increases by 30%. Based on this information, the path planning algorithm generates three candidate paths:

[0052] 1. Take the secondary road to reach Parking Lot B. The estimated driving time is 18 minutes, and the walk to the destination is 5 minutes.

[0053] 2. Detour to Parking Lot C. Estimated driving time: 22 minutes, and an 8-minute walk to the destination.

[0054] 3. Wait for the main road congestion to ease before arriving at Parking Lot A. The estimated total travel time is 25 minutes, and the walk to the destination is 2 minutes.

[0055] Taking into account travel time, walking distance, parking availability and user preferences, these three paths are finally provided to users as candidate parking recommendation path data.

[0056] Step S106: Perform multi-dimensional cross-analysis on the user feedback data and occupancy rate data collected in real time to obtain evaluation index data, wherein the evaluation index data includes: user satisfaction and resource utilization efficiency;

[0057] It should be noted that user feedback data includes user ratings of their parking experience, textual comments, and frequency of use, collected via mobile apps or in-vehicle devices. Occupancy data, derived from each parking lot's real-time monitoring system, reflects parking lot usage. Data preprocessing involves data cleaning, outlier detection, and format standardization. Text and sentiment analysis are required to transform unstructured comments into quantifiable metrics. Occupancy data requires time series analysis to identify usage patterns and trends.

[0058] It uses a variety of data mining and machine learning techniques, such as cluster analysis, association rule mining, and regression analysis, to explore the relationship between user feedback and parking lot occupancy rates. For example, cluster analysis can identify different types of user groups and their parking preferences; association rule mining can discover the correlation between high satisfaction and specific parking lot characteristics (such as location, price, and service quality).

[0059] The calculation of the user satisfaction index takes into account multiple factors, including user ratings, review sentiment, and frequency of use. The resource utilization efficiency index is primarily based on parking lot occupancy data, taking into account factors such as average occupancy, peak usage, and turnover rate. These two indicators are derived through a weighted average or more complex multi-objective evaluation model.

[0060] The generation of evaluation metric data is dynamic and ongoing. The system regularly updates metrics to reflect the latest user feedback and resource utilization. Furthermore, through time series analysis, we can identify changing trends in metrics and help predict future service quality and resource requirements.

[0061] For example: There are three main parking lots A, B, and C in a commercial district. In the past week, the system collected 500 user feedbacks and hourly occupancy data of these three parking lots. After preprocessing, the user feedback data was converted into a satisfaction score of 1-5 and a sentiment score ranging from -1 to 1. The occupancy data was standardized to a scale of 0-100%. Multi-dimensional cross-analysis found that the average satisfaction score of parking lot A was 4.2 and the average sentiment score was 0.6, but its average occupancy rate was only 60%. Parking lot B had a lower score (3.5 points) and a sentiment score of 0.1, but its occupancy rate was as high as 85%. Parking lot C performed more balanced, with a score of 4.0, a sentiment score of 0.4, and an occupancy rate of 75%.

[0062] Through correlation analysis, the system found that high satisfaction is highly correlated with the location convenience and service quality of the parking lot, while high occupancy rate is closely related to price concessions and surrounding commercial activities. Based on these findings, the system calculated the comprehensive user satisfaction index: A is 85 points, B is 70 points, and C is 80 points (out of 100). The resource utilization efficiency index is: A is 65 points, B is 90 points, and C is 85 points. For example, for parking lot A, it is necessary to strengthen marketing strategies to increase occupancy rate; for B, it is necessary to improve service quality to improve user satisfaction; C's performance is relatively balanced, but there is still room for optimization. Through this continuous data analysis and evaluation, the intelligent parking guidance method can continuously optimize its services, balance user experience and resource utilization efficiency, and thus provide higher quality parking guidance services.

[0063] Step S107: Correct the candidate parking suggestion path data using the evaluation index data to obtain target parking suggestion path data, wherein the target parking suggestion path data includes: the geographic coordinates of the target parking lot, the target driving route coordinate sequence, the estimated driving time, the estimated parking fee, and real-time traffic data.

[0064] Specifically, the candidate parking routes are re-evaluated and ranked using previously generated evaluation metrics, including user satisfaction and resource efficiency. This data correction process utilizes a multi-objective optimization algorithm, comprehensively considering multiple factors, including user satisfaction, resource efficiency, travel time, and parking fees. The algorithm assigns different weights to these factors to calculate a comprehensive score for each candidate route. Weights are assigned based on user preferences and system goals, for example, prioritizing user satisfaction or resource efficiency during peak hours. During the correction process, candidate parking lots are first screened. Parking lots with lower evaluation metrics may be eliminated or downgraded. The remaining candidate routes are then optimized. This may include adjusting routes to avoid congested areas or recommending parking lots with higher satisfaction but slightly farther away. The optimization algorithm balances increased travel time with improved parking experience to find the optimal balance.

[0065] Multiple data sources are integrated when generating target parking recommendation path data. The geographic coordinates of the target parking lot are obtained directly from the parking database. The target driving route coordinate sequence is generated using an improved path planning algorithm that takes into account real-time traffic conditions and historical data. The estimated driving time is calculated based on current traffic conditions and historical data and is dynamically updated. The estimated parking fee is inferred from parking lot price data and the estimated parking duration. Real-time traffic data comes from the traffic monitoring system and floating vehicle data, providing real-time traffic conditions for each section of the route. As new evaluation indicator data is generated, the system continuously updates and optimizes parking recommendations. This ensures timely and accurate recommendations that can adapt to rapidly changing traffic and parking situations.

[0066] For example, suppose a user is heading to the downtown business district. The system originally generated three candidate routes, recommending parking lots A, B, and C. Evaluation indicator data shows that parking lot A has a user satisfaction score of 85 and a resource utilization efficiency score of 70; parking lot B has a satisfaction score of 75 and a resource utilization efficiency score of 90; and parking lot C has a satisfaction score of 80 and a resource utilization efficiency score of 85 (all scores are out of 100). Using a multi-objective optimization algorithm, the system re-evaluates these three routes. Assuming the algorithm assigns weights of 0.6 and 0.4 to user satisfaction and resource utilization efficiency, respectively, the combined scores of the three parking lots are: A: 79, B: 81, and C: 82. Based on this result, parking lot C is selected as the target parking lot.

[0067] Next, generate detailed target parking suggestion path data:

[0068] 1. Target parking lot geographic coordinates: longitude 120.5°, latitude 30.3°;

[0069] 2. Target route coordinate sequence: [(120.4°, 30.2°), (120.45°, 30.25°), (120.5°, 30.3°)];

[0070] 3. Estimated travel time: Taking into account real-time traffic conditions, the estimated time is 18 minutes;

[0071] 4. Estimated parking fee: Based on the price of Parking Lot C (10 yuan per hour) and the expected length of stay (2 hours), the estimated parking fee is 20 yuan.

[0072] 5. Real-time traffic data: The main roads along the route are clear, but there is slight congestion near the destination, which is expected to increase the travel time by 2-3 minutes.

[0073] In this embodiment, multi-source traffic data is collected and standardized in real time to generate a target traffic dataset containing traffic flow, parking space availability, environmental factors, and historical trends. This not only improves data timeliness but also enhances decision-making reliability. Time series analysis and pattern recognition are performed on the target traffic dataset to generate traffic flow data and parking demand forecast data, enabling prediction of future traffic conditions and parking demand, thereby enabling preemptive and appropriate resource allocation. Multi-objective optimization is performed on parking demand forecast data based on traffic flow data to generate target decision data, including traffic flow allocation plans, parking resource allocation strategies, and vehicle route recommendations. This target decision data is then converted into multi-channel information formats and prioritized to generate real-time information data suitable for different publishing platforms. This significantly improves the efficiency and reach of information dissemination, enabling users to access parking guidance information across various devices and platforms. Real-time user behavior data, real-time traffic conditions data, and parking lot occupancy data are collected, and correlation analysis and route planning are performed to generate candidate parking recommendation routes, including target parking lots and navigation routes. This achieves precise response to user needs while taking into account real-time road conditions, improving the accuracy and practicality of recommendations. Multi-dimensional cross-analysis and processing are performed on the user feedback data and occupancy data collected in real time to obtain evaluation index data including user satisfaction and resource utilization efficiency. The candidate parking recommendation path data is corrected based on the evaluation index data to obtain the target parking lot geographical coordinates, target driving route coordinate sequence, estimated driving time, estimated parking fee and real-time road condition data. Through multi-step data processing and decision optimization, the entire process from data collection to end-user recommendation is intelligentized, which not only greatly improves parking efficiency, reduces the time and cost of users looking for parking spaces, but also improves the timeliness and accuracy of intelligent parking guidance.

[0074] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0075] (1) Dynamic threshold segmentation is performed on road vehicle density data to obtain traffic flow classification data reflecting traffic flow;

[0076] (2) Perform difference calculation on the vehicle passing data at the parking lot entrance and exit to obtain real-time parking space occupancy rate data;

[0077] (3) Perform semantic analysis on weather condition data and special event information to obtain environmental impact factor data;

[0078] (4) Perform periodic decomposition processing on historical parking data to obtain parking trend cycle data;

[0079] (5) Perform spatiotemporal correlation analysis on traffic flow classification data and real-time parking space occupancy rate data to obtain parking demand hotspot distribution data;

[0080] (6) Perform weighted fusion processing on environmental impact factor data and parking trend cycle data to obtain basic data for parking behavior prediction;

[0081] (7) Perform density clustering on the parking demand hotspot distribution data to obtain regional parking pressure assessment data;

[0082] (8) Perform time series interpolation processing on the basic data of parking behavior prediction to obtain parking demand prediction data in the continuous time domain;

[0083] (9) Adaptively weighting the regional parking pressure assessment data and parking demand forecast data to obtain comprehensive parking index data;

[0084] (10) The comprehensive parking index data is normalized at multiple scales to obtain a target traffic dataset that includes traffic flow, parking space availability, environmental factors, and historical trends.

[0085] Specifically, dynamic threshold segmentation is performed on the road vehicle density data. This step uses a variant of the Otsu method to dynamically adjust the segmentation threshold based on real-time traffic conditions. The calculation formula is as follows:

[0086]

[0087] Among them, T* is the optimal threshold, The traffic flow classification data obtained in this way can more accurately reflect the current traffic conditions. For example, traffic flow can be divided into three levels: low, medium, and high.

[0088] For the vehicle passing data at the entrance and exit of the parking lot, a simple difference calculation is used:

[0089]

[0090] Among them, O r is the occupancy rate, N in is the number of vehicles entering, N out is the number of vehicles leaving, N total The semantic analysis of weather conditions and special event information uses the TF-IDF algorithm:

[0091] TF-IDF(t,d,D)=TF(t,d)×IDF(t,D);

[0092] Here t is a term, d is a document, and D is a document set. The environmental impact factor data obtained in this way can quantify the impact of different events.

[0093] Periodic decomposition of historical parking data using Fast Fourier Transform (FFT):

[0094]

[0095] X k is the frequency domain data, x n is the time domain data, and N is the number of data points. The parking trend cycle data obtained in this way helps to predict future parking demand patterns.

[0096] The spatiotemporal correlation analysis of traffic flow classification data and real-time parking space occupancy data uses the Kriging interpolation method:

[0097]

[0098] Z * (x0) is the value of the predicted point, λ i is the weight, Z(x i ) is the value of the known point. This step generates the parking demand hotspot distribution data.

[0099] The weighted fusion of environmental impact factor data and parking trend cycle data uses the random forest algorithm, where the decision tree is constructed based on the CART algorithm:

[0100]

[0101] Gini(D) is the Gini index, C k is the k-th sample subset.

[0102] Perform DBSCAN density clustering on parking demand hotspot distribution data:

[0103] N ∈ (p)={q∈D|dist(p, q)≤∈};

[0104] N ∈ (p) is the ∈ neighborhood of point p. This step obtains the regional parking pressure assessment data.

[0105] The basic data for parking behavior prediction is processed through cubic spline interpolation:

[0106] S i (x) = a i (xx i ) 3 +b i (xx i ) 2 +c i (xx i )+d i ;

[0107] S i (x) is the interpolation function of the i-th interval.

[0108] x is the independent variable, representing time; is the x-coordinate of the i-th known data point; a i ,b i ,c i ,d i is the unknown coefficient, which can be solved by known data points and continuity conditions.

[0109] Adaptive weighting is performed on regional parking pressure assessment data and parking demand forecast data using the gradient descent method:

[0110]

[0111] w j is the weight, α is the learning rate, and J(w) is the cost function.

[0112] Finally, the comprehensive parking index data is normalized using Min-Max:

[0113]

[0114] x norm is the normalized value, x is the original value, and x min and x max are the minimum and maximum values ​​respectively.

[0115] For example, at 2 p.m. on Saturday in a commercial district, the road density is 80 vehicles per kilometer. Using dynamic threshold segmentation, the threshold is calculated to be 60 vehicles per kilometer, thus classifying it as "high traffic." Parking lot A has 200 vehicles passing through the entrance and 150 vehicles passing through the exit, with a total of 300 parking spaces. The occupancy rate is calculated as (200 - 150) / 300 ≈ 16.7%. The weather forecast of "light rain" is processed using TF-IDF and assigned a weight of 0.6. FFT decomposition reveals the peak period of parking demand during this time. Cubic spline interpolation generates a heat map, showing that demand is highest in the center of the commercial district. Random forest predicts a 20% increase in parking demand during this time. DBSCAN clustering shows that parking pressure in the center of the commercial district reaches 90% saturation. Cubic spline interpolation predicts a continued increase in demand over the next three hours. Gradient descent yields a comprehensive parking index of 85 (out of 100). Finally, Min-Max normalization is performed to the range of 0-1 to form the final target traffic dataset.

[0116] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0117] (1) Perform time window segmentation processing on the target traffic data set to obtain parking data segments of multiple time granularities, and perform Fourier transform processing on the parking data segments of multiple time granularities to obtain parking periodicity feature data;

[0118] (2) Perform wavelet decomposition on parking periodicity characteristic data to obtain multi-scale parking trend data, and perform time series prediction on the multi-scale parking trend data using a long short-term memory network algorithm to obtain initial parking demand forecast data;

[0119] (3) Performing correlation analysis on the initial parking demand forecast data and the real-time traffic flow data collected in real time to obtain the flow-demand correlation factor, and dynamically assigning weights to the flow-demand correlation factor to obtain the parking demand adjustment parameter;

[0120] (4) Adaptively integrate the initial parking demand forecast data and parking demand adjustment parameters to obtain optimized parking demand forecast data, and perform trend extraction and seasonal decomposition on historical traffic flow data to obtain basic traffic flow characteristic data;

[0121] (5) Perform anomaly detection on the basic characteristic data of traffic flow and the real-time traffic event data to obtain the traffic flow fluctuation factor, and perform combined prediction processing on the basic characteristic data of traffic flow and the traffic flow fluctuation factor to obtain traffic flow prediction data and parking demand prediction data.

[0122] Specifically, the target traffic dataset is segmented into time windows to obtain parking data segments at multiple time granularities. This step uses a sliding window technique to segment the continuous time series data into data segments of varying lengths, such as hourly, daily, and weekly. Fourier transforms are performed on these data segments to obtain parking periodicity feature data. The Fourier transform formula is as follows:

[0123]

[0124] Where X(k) is the frequency domain data, x(n) is the time domain data, M is the number of data points, and z is the frequency index.

[0125] Wavelet decomposition is performed on the parking periodicity characteristic data to obtain multi-scale parking trend data. Wavelet decomposition can capture the characteristics of data at different scales. Then, a long short-term memory network (LSTM) algorithm is used to perform time series prediction on the multi-scale parking trend data to obtain initial parking demand forecast data. One of the core formulas of LSTM is:

[0126] f t =σ(W f ·[h t-1 , x t ]+bf );

[0127] Among them, f t is the output of the forget gate, σ is the sigmoid function, W f is the weight matrix, h t-1 is the hidden state of the previous moment, x t is the current input, b f is the bias term.

[0128] Perform correlation analysis on the initial parking demand forecast data and the real-time traffic flow data collected in real time to obtain the flow-demand correlation factor. This step uses the Pearson correlation coefficient:

[0129]

[0130] Among them, r is the correlation coefficient, x i and y i are the observed values ​​of the two variables, and Subsequent steps include dynamic weight allocation, adaptive fusion, trend extraction, seasonality decomposition, anomaly detection, and combined forecasting. These steps utilize a variety of statistical and machine learning techniques to ultimately produce optimized traffic flow and parking demand forecasts.

[0131] For example, a smart parking guidance system in a commercial district collected hourly parking data for the past year. First, the data was segmented into windows at three time scales: 1 hour, 24 hours, and 7 days. A Fourier transform of the 24-hour window revealed a significant peak at frequency k = 1 (corresponding to a 24-hour period) with an amplitude of 500, indicating a clear diurnal periodicity. Wavelet decomposition captured parking demand trends at different scales, such as weekday and weekend differences. After training the LSTM model, parking demand was predicted for the next 24 hours, yielding an initial prediction of an average of 180 vehicles per hour. This prediction was correlated with real-time traffic flow data (e.g., 2,000 vehicles per hour on arterial roads), resulting in a correlation coefficient of r = 0.85, indicating a strong correlation. Based on this correlation, the prediction was adjusted, ultimately resulting in an optimized parking demand forecast of an average of 195 vehicles per hour.

[0132] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0133] (1) Carry out network flow model construction on traffic flow forecast data to obtain traffic flow network topology structure, and carry out spatial clustering processing on parking demand forecast data to obtain parking demand hotspot area data;

[0134] (2) Calculate the matching degree between the traffic flow network topology and the parking demand hotspot area data to obtain the initial traffic flow allocation plan, and perform capacity constraint optimization on the initial traffic flow allocation plan to obtain the traffic flow allocation plan;

[0135] (3) Perform differential analysis on existing parking resource data and parking demand forecast data to obtain parking resource gap data, and dynamically allocate the parking resource gap data to obtain the initial parking resource allocation strategy;

[0136] (4) Conflict detection is performed on the initial parking resource allocation strategy to obtain a parking resource allocation strategy;

[0137] (5) Fusing real-time road network status data and traffic flow allocation schemes to obtain multi-dimensional path evaluation indicators;

[0138] (6) Using the particle swarm optimization algorithm to perform path search processing on the multi-dimensional path evaluation index, a candidate vehicle path set is obtained;

[0139] (7) Perform multi-objective weighing on the candidate vehicle path set to obtain vehicle path recommendation data, and integrate the traffic flow allocation plan, parking resource allocation strategy and vehicle path recommendation data into target decision data.

[0140] Specifically, a network flow model is constructed on the traffic flow forecast data to obtain the traffic flow network topology. The road network is abstracted into a directed graph, where nodes represent intersections, edges represent road segments, and edge weights represent predicted traffic flows. Simultaneously, spatial clustering is performed on the parking demand forecast data to obtain parking demand hotspots. DBSCAN (density-based spatial clustering algorithm) is used to identify areas with high parking demand.

[0141] The matching degree between the traffic network topology and parking demand hotspot data is calculated to generate an initial traffic flow allocation plan. This matching degree calculation takes into account road capacity, predicted traffic volume, and surrounding parking demand, using a weighted matching algorithm to allocate traffic flow. The initial traffic flow allocation plan is then optimized under capacity constraints to generate the final traffic flow allocation plan. This step uses the Frank-Wolfe algorithm to minimize overall travel time while satisfying road capacity constraints. Regarding parking resource allocation, a discrepancy analysis is first performed between existing parking resource data and predicted parking demand data to generate parking resource gap data. This discrepancy analysis uses a simple supply-demand comparison to identify areas with insufficient parking spaces. The parking resource gap data is then dynamically allocated to generate an initial parking resource allocation strategy. This step uses a greedy algorithm to prioritize resources to areas with the largest gaps. Conflict detection is then performed on the initial parking resource allocation strategy to generate the final parking resource allocation strategy. Conflict detection primarily addresses issues such as duplicate and uneven resource allocation, resolving conflicts through iterative adjustments.

[0142] To generate vehicle route recommendation data, real-time road network status data and traffic flow allocation plans are first integrated to obtain multidimensional route evaluation metrics. These metrics include expected travel time, route length, congestion level, and parking convenience. A particle swarm optimization algorithm is then used to perform a path search on these multidimensional route evaluation metrics, generating a set of candidate vehicle routes. The particle swarm optimization algorithm simulates swarm intelligence and can find near-optimal solutions in complex route spaces. Finally, a multi-objective trade-off is performed on the candidate vehicle routes to generate vehicle route recommendation data. This multi-objective trade-off utilizes the Pareto optimality principle to balance factors such as travel time, distance, and parking convenience. Ultimately, the traffic flow allocation plan, parking resource allocation strategy, and vehicle route recommendation data are integrated into target decision data, providing comprehensive decision support for the intelligent parking guidance system.

[0143] For example, a city's commercial district faced severe traffic congestion and parking difficulties on Saturday afternoons. Using a network flow model, the area was abstracted into a directed graph consisting of 50 nodes (intersections) and 80 edges (road segments). The DBSCAN clustering algorithm identified three major parking demand hotspots: located near shopping malls, food courts, and entertainment venues. An initial traffic flow allocation plan showed that traffic on the main arterial road exceeded 120% of its capacity. After optimization using the Frank-Wolfe algorithm, 20% of traffic was allocated to parallel secondary arterial roads, reducing traffic on the main arterial road to 105% of its capacity. Parking resource analysis revealed that the shopping mall area had the largest parking shortage, reaching 200 spaces. A dynamic allocation strategy determined that 150 temporary parking spaces would be opened in a nearby office building and 50 additional on-street temporary parking spaces would be added within a 500-meter radius. A particle swarm optimization algorithm generated 3-5 candidate routes for each vehicle, taking into account travel time, distance, and parking convenience. For example, for a vehicle whose destination is a shopping mall, the algorithm recommends a route that detours around a secondary road. Although the distance increases by 10%, it is expected to save 15 minutes of driving time and has an 80% probability of finding a parking space.

[0144] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0145] (1) Perform data compression on the traffic flow allocation scheme to obtain traffic flow summary data;

[0146] (2) Perform spatial index construction on parking resource allocation strategies to obtain parking space distribution data;

[0147] (3) Simplify the vehicle path recommendation data to obtain key path node data;

[0148] (4) Perform data fusion processing on traffic flow summary data, parking space distribution data, and key path node data to obtain information data packets;

[0149] (5) Perform multi-level cache strategy processing on the information data packet to obtain a hierarchical storage information cache structure, and perform real-time update frequency analysis on the information cache structure to obtain information timeliness score data;

[0150] (6) Prioritizing the information timeliness score data to obtain an information release queue, and performing multi-channel adaptation processing on the data in the information release queue to obtain formatted information for different release platforms;

[0151] (7) Securely encrypt the formatted information for different publishing platforms to obtain real-time information data suitable for different publishing platforms.

[0152] Specifically, the traffic flow allocation plan is compressed to obtain traffic flow summary data. This step uses the discrete cosine transform (DCT) compression algorithm to convert the original traffic flow allocation data into a frequency domain representation and retain the main frequency components, thereby significantly reducing the amount of data but retaining key information. The parking resource allocation strategy is spatially indexed to obtain parking space distribution data. Here, the R-tree index structure is used to organize the parking space location information in the two-dimensional space into a hierarchical tree structure for fast query and retrieval. For the vehicle path recommendation data, path simplification is performed to obtain key path node data. This step applies the Douglas-Peucker algorithm to simplify the path by gradually deleting non-critical points while maintaining the overall shape of the path and key turning points.

[0153] Traffic flow summary data, parking space distribution data, and key path node data are fused to generate an information data packet. This fusion process utilizes a multi-level data structure to organize different types of data into a unified format for easy subsequent processing and transmission. A multi-level caching strategy is applied to the information data packet, resulting in a hierarchical information cache structure. This step utilizes the LRU (least recently used) caching algorithm to separate data into hot and cold data, with hot data stored in the fast-access cache and cold data in the slower storage tier. The information cache structure is analyzed for real-time update frequency to generate information timeliness scores. A sliding window technique is employed to calculate the access frequency of each piece of information within a recent time window and assign a timeliness score based on this frequency. The information timeliness score data is then prioritized to generate an information release queue. This sorting utilizes a heap sort algorithm to ensure that highly time-sensitive information is released first.

[0154] Data in the information publishing queue is adapted for multiple channels to generate formatted information for different publishing platforms. This step uses the adapter pattern to customize the data format and display method for each publishing platform (such as mobile apps, in-vehicle devices, roadside displays, etc.). Finally, the formatted information for each publishing platform is securely encrypted to generate real-time information data suitable for each publishing platform. The encryption process uses the AES (Advanced Encryption Standard) algorithm to ensure data security during transmission.

[0155] For example, an intelligent parking guidance system in a city's commercial district processed a large amount of real-time data on a Saturday afternoon. The original traffic flow distribution plan contained 5,000 data points. After DCT compression, 500 key frequency components were retained, reducing the data volume by 90% while retaining 95% of the information. There are 200 parking lots in the area. Using an R-tree index, query time was reduced from linear O(n) to logarithmic O(log n). A recommended route originally consisted of 100 GPS coordinates. After traffic trajectory thinning (Douglas-Peucker) algorithm, 20 key nodes were retained, and the path length error was less than 2%.

[0156] The fused data packet is 2MB in size. A multi-level caching strategy is used to store 500KB of hotspot data in memory, with the remainder stored on a solid-state drive. Real-time update frequency analysis shows that parking information is accessed three times more frequently than traffic flow information, thus receiving higher priority in the information release queue. For mobile app users, a 50KB JSON-formatted data packet is generated, while for roadside displays, a simplified text format of only 10KB is generated. All data is encrypted with AES-256 before transmission to ensure security.

[0157] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0158] (1) Perform cluster analysis on user parking behavior data to obtain user parking preference information, and calculate road congestion degree on traffic condition data to obtain traffic flow index;

[0159] (2) Perform difference statistical processing on parking lot occupancy rate data to obtain dynamic parking space occupancy rate;

[0160] (3) Perform correlation analysis on the user's parking preference signal and the dynamic parking space occupancy rate to obtain a personalized parking lot recommendation list;

[0161] (4) Perform threshold segmentation on the traffic flow index to obtain a set of feasible driving sections;

[0162] (5) Perform path planning on the set of feasible driving sections through a heuristic search algorithm to obtain multiple candidate navigation routes;

[0163] (6) Extracting geographic coordinates of parking lots in the personalized parking lot recommendation list to obtain a target parking lot coordinate set;

[0164] (7) Calculate the matching degree of multiple candidate navigation routes and target parking lot coordinate sets to obtain a route-parking lot matching solution;

[0165] (8) The route-parking lot matching scheme is comprehensively scored to obtain the optimal parking recommendation path, and the optimal parking recommendation path is data encapsulated to obtain candidate parking recommendation path data including the target parking lot and the navigation route.

[0166] Specifically, cluster analysis is performed on user parking behavior data to obtain user parking preference information. This step uses the K-means clustering algorithm to divide users into different groups based on their historical parking preferences (such as distance to destination, price, parking lot type, etc.). The objective function of the K-means algorithm can be expressed as:

[0167]

[0168] Among them, J is the objective function, h is the number of clusters, and n is the number of data points. is the i-th data point in the j-th class, c j is the cluster center of the jth class.

[0169] Traffic congestion data is processed to calculate the traffic flow index, which is expressed as the ratio of the average vehicle speed to the road design speed:

[0170]

[0171] F i is the fluency index, V a is the actual average speed, V d Design speed for the road.

[0172] Parking lot occupancy data is statistically processed using differences to generate dynamic parking space occupancy rates. Correlation analysis is then performed between user parking preferences and dynamic parking space occupancy rates to generate a personalized parking recommendation list. This step utilizes a weighted scoring system that comprehensively considers user preferences and current parking conditions. A threshold segmentation process is applied to the traffic flow index to generate a set of feasible driving segments. Typically, segments with a flow index greater than 70% are considered feasible.

[0173] By using a heuristic search algorithm (such as the A* algorithm) to perform path planning on a set of feasible driving sections, multiple candidate navigation routes are obtained. The evaluation function of the A* algorithm is:

[0174] f(n)=g(n)+h(n);

[0175] f(n) is the estimated total cost of node n, g(n) is the actual cost from the starting point to n, and h(n) is the estimated cost from n to the end point.

[0176] The geographic coordinates of the parking lots in the personalized parking lot recommendation list are extracted to obtain the target parking lot coordinate set. A matching degree is then calculated between multiple candidate navigation routes and the target parking lot coordinate set to obtain a route-parking lot matching solution. This matching degree calculation takes into account factors such as the distance between the route endpoint and the parking lot and the estimated travel time.

[0177] Finally, the route-parking lot matching scheme is comprehensively scored to obtain the optimal parking recommendation path, and the optimal parking recommendation path is data encapsulated to obtain the candidate parking recommendation path data including the target parking lot and the navigation route.

[0178] For example, a user is heading to a shopping district on Saturday afternoon. The system analyzes the user's historical parking data and finds that they prefer reasonably priced parking within 500 meters of their destination. Using the K-means algorithm (k = 3), the user is classified as belonging to the "convenience-first" group. Real-time traffic data shows that the average speed on Main Road A is 40 km / h, while the design speed is 60 km / h. The calculated smoothness index FA = 66.7%, below the 70% threshold, excludes it from the feasible driving section. Secondary Road B, with a smoothness index FB = 85%, is included in the feasible section.

[0179] Three parking lots, P1, P2, and P3, are detected near a commercial area, with current occupancy rates of 85%, 60%, and 40%, respectively. Taking into account user preferences and current occupancy, the system generates a personalized parking recommendation list for the user: [P2, P3, P1]. Using the A* algorithm, three candidate routes are planned along the feasible road segment. Route R1, which reaches P2, has the lowest f(n) value and is the optimal choice. After comprehensive evaluation, the system recommends that the user choose parking lot P2, reached via route R1. This recommended route has an estimated travel time of 15 minutes, is 450 meters from the destination, and costs 10 yuan per hour. The system packages this information into a data package containing the coordinates of the target parking lot P2 (longitude: 120.5°, latitude: 30.3°) and the coordinate sequence of key nodes in navigation route R1, and provides it to the user as the final candidate parking route data.

[0180] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0181] (1) Perform sentiment analysis on user feedback data to obtain an initial parking experience satisfaction score, and then perform time-attenuation processing on the initial parking experience satisfaction score to obtain time-weighted satisfaction data;

[0182] (2) Perform time series decomposition processing on parking lot occupancy rate data to obtain parking lot utilization efficiency trend data, and perform correlation analysis on the timeliness weighted satisfaction data and parking lot utilization efficiency trend data to obtain the satisfaction-efficiency correlation index;

[0183] (3) Extracting and processing the distribution characteristics of user parking time data to obtain parking time pattern data, and calculating the matching degree between parking time pattern data and parking lot turnover rate data to obtain the time-space utilization efficiency index;

[0184] (4) The satisfaction-efficiency correlation index and the time-space utilization efficiency index are integrated to obtain the initial value of the comprehensive evaluation, and the parking guidance response time data are statistically analyzed to obtain the response efficiency score;

[0185] (5) Perform weighted average processing on the initial value of comprehensive evaluation and response efficiency score to obtain the original data of target evaluation indicators;

[0186] (6) Normalizing and quantifying the original data of the target evaluation indicators to obtain evaluation indicator data, wherein the evaluation indicator data includes: user satisfaction and resource utilization efficiency.

[0187] Specifically, the evaluation process for intelligent parking guidance involves multiple complex data processing steps designed to comprehensively assess system performance and user experience. First, sentiment analysis is performed on user feedback data to generate an initial parking experience satisfaction score. This step uses natural language processing technology to convert text reviews into numerical scores. The core formula for sentiment analysis is:

[0188]

[0189] Among them, S is the sentiment score, w i is the weight of the i-th word, p i is the sentiment polarity value of the i-th word, and n is the number of words in the review. The initial parking experience satisfaction score is subjected to time decay processing to obtain time-weighted satisfaction data. The time decay function uses an exponential decay model:

[0190] W(t)=S·e -λt

[0191] W(t) is the weighted satisfaction after time t, λ is the decay coefficient, and t is the time interval.

[0192] Parking lot occupancy data is decomposed into time series to generate parking lot utilization efficiency trend data. The STL (Seasonal and Trend decomposition using Loess) method is used to decompose the time series into trend, seasonal, and residual components. Next, correlation analysis is performed on the timeliness-weighted satisfaction data and parking lot utilization efficiency trend data to generate a satisfaction-efficiency correlation index. This is calculated using the Pearson correlation coefficient.

[0193] The distribution characteristics of user parking time data are extracted to obtain parking time pattern data. Kernel density estimation is used to identify the distribution characteristics of parking time. The parking time pattern data and parking lot turnover rate data are then matched to obtain a spatiotemporal efficiency index. This matching calculation considers the degree of fit between the parking time distribution and the parking lot's designed turnover rate. The satisfaction-efficiency correlation index and the spatiotemporal efficiency index are fused to obtain the initial value for the comprehensive evaluation. This fusion process utilizes a weighted summation method. Simultaneously, statistical analysis is performed on parking guidance response time data to obtain a response efficiency score. The cumulative distribution function (CDF) is used to evaluate response time performance. A weighted average of the initial value for the comprehensive evaluation and the response efficiency score is performed to obtain the raw evaluation index data. This data is then normalized and quantified to obtain the final evaluation index data, including user satisfaction and resource utilization efficiency. Normalization utilizes the Min-Max method to map the data to the [0, 1] interval.

[0194] For example, a commercial parking lot collected 1,000 user feedback items over the past month. Through sentiment analysis, the system calculated an initial average satisfaction rating of 4.2 out of 5. Applying a time decay function (λ = 0.05), the weight of ratings from a week ago dropped to approximately 70%. STL decomposition of parking lot occupancy data revealed that weekend utilization was 20% higher than weekday utilization. The calculated satisfaction-efficiency correlation index was 0.75, indicating a strong positive correlation.

[0195] Analysis of parking duration data revealed an average parking time of 2.5 hours, which closely matches the designed turnover rate of six times per day, resulting in a time-space efficiency index of 0.85. The initial value of the comprehensive evaluation was calculated to be 0.80. The parking guidance system's average response time was 3 seconds, with 90% of requests being responded to within 5 seconds, resulting in a response efficiency score of 0.88. Finally, normalization of the raw evaluation metrics yielded a user satisfaction index of 0.85 and a resource utilization efficiency index of 0.82.

[0196] In a specific embodiment, the process of executing step S107 may specifically include the following steps:

[0197] (1) Perform threshold analysis on the user satisfaction in the evaluation index data to obtain a parking lot satisfaction ranking list;

[0198] (2) Extracting geographic information from the target parking lot in the candidate parking suggestion path data to obtain a candidate parking lot coordinate set, and matching and screening the parking lot satisfaction ranking list with the candidate parking lot coordinate set to obtain the geographic coordinates of the target parking lot;

[0199] (3) performing path smoothing processing on the navigation routes in the candidate parking suggestion route data to obtain an initial driving route coordinate sequence, and optimizing the initial driving route coordinate sequence to obtain a target driving route coordinate sequence;

[0200] (4) Segment the target route coordinate sequence into sections to obtain a set of section units, and then fuse the section unit set with real-time traffic flow data to obtain the estimated travel time for each section;

[0201] (5) Accumulate and calculate the estimated travel time of each road section to obtain the estimated travel time;

[0202] (6) Perform time series forecasting on the historical price data of the target parking lot to obtain the estimated parking fee;

[0203] (7) The collected road condition data is semantically processed to obtain real-time road condition data, and the geographical coordinates of the target parking lot, the target driving route coordinate sequence, the estimated driving time, the estimated parking fee and the real-time road condition data are integrated into the target parking recommended path data.

[0204] Specifically, a threshold analysis is performed on the user satisfaction scores in the evaluation index data to generate a ranked parking lot satisfaction list. Using the quantile method, parking lots with satisfaction scores above the 75th percentile are selected and ranked from high to low. Next, geographic information is extracted from the target parking lot in the candidate parking suggestion route data to generate a set of candidate parking lot coordinates. Geocoding technology is used here to convert parking lot addresses into longitude and latitude coordinates. The ranked parking lot satisfaction list and the candidate parking lot coordinate set are then matched and screened to obtain the geographic coordinates of the target parking lot. This matching process considers both satisfaction ranking and geographic location to select the optimal parking lot.

[0205] The navigation routes in the candidate parking suggestion path data are smoothed to obtain an initial route coordinate sequence. This step uses the Douglas-Peucker algorithm to simplify redundant points in the path while retaining key turning points. The initial route coordinate sequence is then optimized to obtain a target route coordinate sequence. This optimization process takes into account road conditions, such as one-way streets and turn restrictions, and uses the A* algorithm for path replanning. The target route coordinate sequence is segmented to obtain a set of segment units. This segmentation is based on intersections and major turning points, dividing the entire route into multiple independent segments. This set of segment units is then fused and analyzed with real-time traffic flow data to obtain the estimated travel time for each segment. This step uses historical traffic data and real-time traffic information to calculate the estimated travel time for each segment using a weighted average method. The estimated travel time for each segment is then accumulated to obtain the overall estimated travel time.

[0206] The target parking lot's historical price data is processed through time series forecasting to obtain estimated parking fees. The ARIMA (Autoregressive Integrated Moving Average) model is used here to predict parking fees for a certain period of time in the future, taking into account factors such as time, date, and seasonality. Simultaneously, the collected traffic data is semantically processed to obtain real-time traffic data. Semantic processing converts digitized traffic information into user-friendly text descriptions, such as "unimpeded" and "mildly congested." The target parking lot's geographic coordinates, the target driving route's coordinate sequence, the estimated travel time, the estimated parking fee, and real-time traffic data are integrated into the target parking route's recommended path data. This integration process creates a structured data object containing all necessary information for subsequent display and use.

[0207] For example, a user uses the intelligent parking guidance system on a Saturday afternoon to find a parking space near a commercial district. First, user satisfaction data for 10 nearby parking lots is analyzed, with a satisfaction threshold of 4.2 points (out of a 5-point scale) set to identify five parking lots with high satisfaction. After extracting geographic information, the coordinates of these five parking lots are mapped onto a two-dimensional plane. The matching and selection process takes into account satisfaction rankings and proximity to the destination, ultimately selecting Parking Lot A with coordinates (39.9153, 116.4038) as the target parking lot.

[0208] The initial navigation route consisted of 100 coordinate points, which were smoothed using the Douglas-Peucker algorithm, retaining 25 key points. The A* algorithm optimized the route, taking into account actual road conditions, such as one-way restrictions, to generate the final target route coordinate sequence. The route was divided into eight segments, and the estimated travel time for each segment was calculated using real-time traffic flow data. For example, the first segment was estimated to take 3 minutes, the second 5 minutes, and so on. The cumulative calculation yielded a total estimated travel time of 22 minutes. An ARIMA model analyzed parking lot A's price data from the past three months and predicted that the parking fee for that afternoon would be 15 yuan per hour. Real-time traffic data, after semantic processing, showed that 70% of the route segments were "unblocked," 20% were "lightly congested," and 10% were "moderately congested." Finally, the generated target parking recommendation path data includes: the coordinates of the target parking lot A (39.9153, 116.4038), a driving route coordinate sequence consisting of 25 key points, an estimated driving time of 22 minutes, an estimated parking fee of 15 yuan / hour, and detailed real-time traffic information.

[0209] The above describes the intelligent parking guidance method in the embodiment of the present application. The following describes the intelligent parking guidance system in the embodiment of the present application. Figure 2 In the embodiment of the present application, an embodiment of the intelligent parking guidance system includes:

[0210] Processing module 201 is used to collect and standardize multi-source traffic data in real time to obtain a target traffic dataset containing traffic flow, parking space availability, environmental factors and historical trends;

[0211] Identification module 202, configured to perform time series analysis and pattern recognition processing on the target traffic data set to obtain traffic flow data and parking demand prediction data;

[0212] An optimization module 203 is configured to perform multi-objective optimization processing on the parking demand prediction data based on the vehicle flow data to obtain target decision data, wherein the target decision data includes: a traffic flow allocation plan, a parking resource allocation strategy, and vehicle path recommendation data;

[0213] The conversion module 204 is used to convert the target decision data into a multi-channel information format and perform priority sorting processing to obtain real-time information data suitable for different publishing platforms;

[0214] Planning module 205 is configured to collect user parking behavior data, real-time traffic status data, and parking lot occupancy rate data in real time, and perform correlation analysis and path planning on the user parking behavior data, the real-time traffic status data, and the parking lot occupancy rate data to obtain candidate parking path data including a target parking lot and a navigation route;

[0215] The analysis module 206 is configured to perform multi-dimensional cross-analysis processing on the user feedback data and the occupancy rate data collected in real time to obtain evaluation index data, wherein the evaluation index data includes user satisfaction and resource utilization efficiency;

[0216] The correction module 207 is used to correct the candidate parking path data using the evaluation index data to obtain target parking path data, wherein the target parking path data includes: the geographic coordinates of the target parking lot, the target driving route coordinate sequence, the estimated driving time, the estimated parking fee, and real-time traffic data.

[0217] Through the collaborative efforts of these components, multi-source traffic data is collected and standardized in real time to generate a target traffic dataset encompassing traffic flow, parking availability, environmental factors, and historical trends. This not only improves data timeliness but also enhances decision-making reliability. Time series analysis and pattern recognition are performed on the target traffic dataset to generate traffic flow data and parking demand forecasts, enabling prediction of future traffic conditions and parking demand, enabling proactive resource allocation. Multi-objective optimization is performed on parking demand forecasts based on traffic flow data to generate target decision data, including traffic flow allocation plans, parking resource allocation strategies, and vehicle route recommendations. This target decision data is then converted into multi-channel information formats and prioritized to generate real-time information suitable for different distribution platforms. This significantly improves the efficiency and reach of information dissemination, enabling users to access parking guidance information across various devices and platforms. Real-time user behavior data, traffic conditions data, and parking lot occupancy data are collected, and correlation analysis and route planning are performed to generate candidate parking recommendation routes, including target parking lots and navigation routes. This precisely responds to user needs while also taking into account real-time road conditions, improving the accuracy and practicality of recommendations. Multi-dimensional cross-analysis and processing are performed on the user feedback data and occupancy data collected in real time to obtain evaluation index data including user satisfaction and resource utilization efficiency. The candidate parking recommendation path data is corrected based on the evaluation index data to obtain the target parking lot geographical coordinates, target driving route coordinate sequence, estimated driving time, estimated parking fee and real-time road condition data. Through multi-step data processing and decision optimization, the entire process from data collection to end-user recommendation is intelligentized, which not only greatly improves parking efficiency, reduces the time and cost of users looking for parking spaces, but also improves the timeliness and accuracy of intelligent parking guidance.

[0218] The present application also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, which, when executed on a computer, cause the computer to execute the steps of the intelligent parking guidance method.

[0219] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0220] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An intelligent parking guidance method, characterized in that: The intelligent parking guidance method comprises: Real-time collection and standardization of multi-source traffic data to obtain a target traffic dataset that includes traffic flow, parking availability, environmental factors, and historical trends; The target traffic data set is subjected to time series analysis and pattern recognition processing to obtain traffic flow prediction data and parking demand prediction data, including: performing time window segmentation processing on the target traffic data set to obtain parking data segments of multiple time granularities, and performing Fourier transform processing on the parking data segments of multiple time granularities to obtain parking periodic feature data; performing wavelet decomposition processing on the parking periodic feature data to obtain multi-scale parking trend data, and performing time series prediction processing on the multi-scale parking trend data through a long short-term memory network algorithm to obtain initial parking demand prediction data; performing correlation analysis processing on the initial parking demand prediction data and real-time traffic flow data collected in real time to obtain a flow-demand correlation factor, and Dynamic weight allocation processing is performed on the traffic-demand correlation factor to obtain a parking demand adjustment parameter; the initial parking demand forecast data and the parking demand adjustment parameter are adaptively fused to obtain optimized parking demand forecast data, and trend extraction and seasonal decomposition processing are performed on historical traffic flow data to obtain basic traffic flow characteristic data; anomaly detection processing is performed on the basic traffic flow characteristic data and real-time traffic event data to obtain a traffic flow fluctuation factor, and a combined prediction processing is performed on the basic traffic flow characteristic data and the traffic flow fluctuation factor to obtain the traffic flow forecast data and the parking demand forecast data; in addition to considering the time trend of historical parking data, parking demand forecasting also needs to analyze the correlation between parking behavior, weather, and surrounding activities; Based on the traffic flow prediction data, performing multi-objective optimization processing on the parking demand prediction data to obtain target decision data; Perform multi-channel information format conversion and priority sorting on the target decision data to obtain real-time information data suitable for different publishing platforms; Collecting user parking behavior data, real-time traffic status data, and updated parking lot occupancy rate data in real time, and performing correlation analysis and path planning on the user parking behavior data, the real-time traffic status data, and the parking lot occupancy rate data to obtain candidate parking recommendation path data including a target parking lot and a navigation route; Performing multi-dimensional cross-analysis processing on the user feedback data and the occupancy rate data collected in real time to obtain evaluation index data; The candidate parking suggestion path data is corrected using the evaluation index data to obtain target parking suggestion path data.

2. The intelligent parking guidance method according to claim 1, characterized in that: The multi-source traffic data includes road vehicle density data, parking lot entrance and exit vehicle traffic data, weather condition data, special event information, and historical parking data. The multi-source traffic data is collected and standardized in real time to obtain a target traffic dataset containing traffic flow, parking space availability, environmental factors, and historical trends, including: Performing dynamic threshold segmentation processing on the road vehicle density data to obtain vehicle flow classification data reflecting traffic flow; Performing difference calculation on the vehicle passing data at the parking lot entrance and exit to obtain real-time parking space occupancy rate data; Performing semantic analysis on the weather condition data and the special event information to obtain environmental impact factor data; Performing periodic decomposition processing on the historical parking data to obtain parking trend period data; Performing spatiotemporal correlation analysis on the traffic flow classification data and the real-time parking space occupancy rate data to obtain parking demand hotspot distribution data; Performing weighted fusion processing on the environmental impact factor data and the parking trend cycle data to obtain basic data for parking behavior prediction; Performing density clustering processing on the parking demand hotspot distribution data to obtain regional parking pressure assessment data; Performing time series interpolation processing on the parking behavior prediction basic data to obtain parking demand prediction data in a continuous time domain; Performing adaptive weight distribution processing on the regional parking pressure assessment data and the parking demand prediction data to obtain comprehensive parking index data; The comprehensive parking index data is subjected to multi-scale normalization processing to obtain a target traffic dataset including traffic flow, parking space availability, environmental factors and historical trends.

3. The intelligent parking guidance method according to claim 1, characterized in that: The parking demand forecast data is subjected to multi-objective optimization processing based on the traffic flow forecast data to obtain target decision data, wherein the target decision data includes: traffic flow allocation plan, parking resource allocation strategy and vehicle path recommendation data, including: Performing network flow model construction processing on the traffic flow prediction data to obtain a traffic flow network topology structure, and performing spatial clustering processing on the parking demand prediction data to obtain parking demand hotspot area data; Performing matching calculation processing on the traffic flow network topology structure and the parking demand hotspot area data to obtain an initial traffic flow allocation plan, and performing capacity constraint optimization processing on the initial traffic flow allocation plan to obtain the traffic flow allocation plan; Performing difference analysis on existing parking resource data and the parking demand forecast data to obtain parking resource gap data, and dynamically allocating the parking resource gap data to obtain an initial parking resource allocation strategy; Performing conflict detection processing on the initial parking resource allocation strategy to obtain the parking resource allocation strategy; The real-time road network status data and the traffic flow allocation scheme are integrated to obtain a multi-dimensional path evaluation index; Performing path search processing on the multi-dimensional path evaluation index by using a particle swarm optimization algorithm to obtain a candidate vehicle path set; A multi-objective trade-off process is performed on the candidate vehicle path set to obtain the vehicle path recommendation data, and the traffic flow allocation plan, the parking resource allocation strategy and the vehicle path recommendation data are integrated into the target decision data.

4. The intelligent parking guidance method according to claim 3, characterized in that: The target decision data is subjected to multi-channel information format conversion and priority sorting processing to obtain real-time information data suitable for different publishing platforms, including: Performing data compression processing on the traffic flow allocation plan to obtain traffic flow summary data; performing spatial index construction processing on the parking resource allocation strategy to obtain parking space distribution data; Performing path simplification processing on the vehicle path recommendation data to obtain key path node data; Performing data fusion processing on the traffic flow summary data, the parking space distribution data, and the key path node data to obtain an information data packet; Performing multi-level cache strategy processing on the information data packet to obtain a hierarchically stored information cache structure, and performing real-time update frequency analysis processing on the information cache structure to obtain information timeliness score data; Prioritizing the information timeliness score data to obtain an information release queue, and performing multi-channel adaptation processing on the data in the information release queue to obtain formatted information for different release platforms; The formatted information for different publishing platforms is securely encrypted to obtain the real-time information data applicable to the different publishing platforms.

5. The intelligent parking guidance method according to claim 1, characterized in that: The real-time collection of user parking behavior data, real-time traffic status data, and parking lot occupancy rate data, and the correlation analysis and path planning processing of the user parking behavior data, the real-time traffic status data, and the parking lot occupancy rate data, to obtain candidate parking recommendation path data including a target parking lot and a navigation route, includes: Performing cluster analysis on the user parking behavior data to obtain user parking preference information, and calculating road congestion on the traffic condition data to obtain a traffic fluency index; Performing difference statistical processing on the parking lot occupancy rate data to obtain a dynamic parking space occupancy rate; Performing correlation analysis on the user parking preference signal and the dynamic parking space occupancy rate to obtain a personalized parking lot recommendation list; Performing threshold segmentation processing on the traffic fluency index to obtain a set of feasible driving sections; Performing path planning on the set of feasible driving sections by a heuristic search algorithm to obtain a plurality of candidate navigation routes; Extracting geographic coordinates of parking lots in the personalized parking lot recommendation list to obtain a target parking lot coordinate set; Performing matching calculation on the plurality of candidate navigation routes and the target parking lot coordinate set to obtain a route-parking lot matching solution; The route-parking lot matching scheme is comprehensively scored to obtain an optimal parking suggestion path, and the optimal parking suggestion path is data-encapsulated to obtain candidate parking suggestion path data including a target parking lot and a navigation route.

6. The intelligent parking guidance method according to claim 5, characterized in that: The user feedback data and the occupancy rate data collected in real time are subjected to multi-dimensional cross-analysis processing to obtain evaluation index data, wherein the evaluation index data includes: user satisfaction and resource utilization efficiency, including: Performing sentiment analysis on the user feedback data to obtain an initial parking experience satisfaction score, and performing time decay processing on the initial parking experience satisfaction score to obtain timeliness-weighted satisfaction data; Performing time series decomposition processing on parking lot occupancy rate data to obtain parking lot utilization efficiency trend data, and performing correlation analysis processing on the timeliness weighted satisfaction data and the parking lot utilization efficiency trend data to obtain a satisfaction-efficiency correlation index; Performing distribution feature extraction on user parking time data to obtain parking time pattern data, and performing matching calculation on the parking time pattern data and parking lot turnover rate data to obtain a time-space utilization efficiency index; The satisfaction-efficiency correlation index and the time-space utilization efficiency index are integrated to obtain an initial value for comprehensive evaluation, and the parking guidance response time data are statistically analyzed to obtain a response efficiency score; Performing weighted averaging processing on the comprehensive evaluation initial value and the response efficiency score to obtain the target evaluation index original data; Normalizing and quantifying the original data of the target evaluation index to obtain the evaluation index data, wherein the evaluation index data includes: user satisfaction and resource utilization efficiency.

7. The intelligent parking guidance method according to claim 6, characterized in that: The candidate parking suggestion path data is corrected using the evaluation index data to obtain target parking suggestion path data, wherein the target parking suggestion path data includes: geographic coordinates of the target parking lot, a target driving route coordinate sequence, an estimated driving time, an estimated parking fee, and real-time traffic data, including: Performing threshold analysis on the user satisfaction in the evaluation index data to obtain a parking lot satisfaction ranking list; Extracting geographic information of a target parking lot from the candidate parking suggestion route data to obtain a candidate parking lot coordinate set, and performing matching and screening processing on the parking lot satisfaction ranking list and the candidate parking lot coordinate set to obtain geographic coordinates of the target parking lot; performing path smoothing processing on the navigation routes in the candidate parking suggestion route data to obtain an initial driving route coordinate sequence, and performing optimization processing on the initial driving route coordinate sequence to obtain a target driving route coordinate sequence; Performing segmentation processing on the target driving route coordinate sequence to obtain a set of road section units, and performing fusion analysis processing on the set of road section units and real-time traffic flow data to obtain an estimated travel time for each road section; and performing cumulative calculation processing on the estimated travel time for each road section to obtain the estimated travel time; Performing time series forecasting on the historical price data of the target parking lot to obtain an estimated parking fee; The collected traffic condition data is semantically processed to obtain the real-time traffic condition data, and the geographical coordinates of the target parking lot, the target driving route coordinate sequence, the estimated driving time, the estimated parking fee and the real-time traffic condition data are integrated into the target parking recommended path data.

8. An intelligent parking guidance system, characterized in that: For executing the intelligent parking guidance method according to any one of claims 1 to 7, the intelligent parking guidance system comprises: A processing module is used to collect and standardize multi-source traffic data in real time to obtain a target traffic dataset containing traffic flow, parking space availability, environmental factors and historical trends; The identification module is used to perform time series analysis and pattern recognition processing on the target traffic data set to obtain vehicle flow data and parking demand prediction data, including: performing time window segmentation processing on the target traffic data set to obtain parking data segments of multiple time granularities, and performing Fourier transform processing on the parking data segments of multiple time granularities to obtain parking periodic feature data; performing wavelet decomposition processing on the parking periodic feature data to obtain multi-scale parking trend data, and performing time series prediction processing on the multi-scale parking trend data through a long short-term memory network algorithm to obtain initial parking demand prediction data; performing correlation analysis processing on the initial parking demand prediction data and real-time traffic flow data collected in real time to obtain a flow-demand correlation factor, and Dynamic weight allocation processing is performed on the traffic-demand correlation factor to obtain a parking demand adjustment parameter; the initial parking demand forecast data and the parking demand adjustment parameter are adaptively fused to obtain optimized parking demand forecast data, and trend extraction and seasonal decomposition processing are performed on historical traffic flow data to obtain basic traffic flow characteristic data; anomaly detection processing is performed on the basic traffic flow characteristic data and real-time traffic event data to obtain a traffic flow fluctuation factor, and a combined prediction processing is performed on the basic traffic flow characteristic data and the traffic flow fluctuation factor to obtain the traffic flow forecast data and the parking demand forecast data; in addition to considering the time trend of historical parking data, parking demand forecasting also requires analyzing the correlation between parking behavior, weather, and surrounding activities; an optimization module, configured to perform multi-objective optimization processing on the parking demand prediction data based on the vehicle flow data to obtain target decision data; A conversion module is used to convert the target decision data into multiple information formats and perform priority sorting processing to obtain real-time information data suitable for different publishing platforms; a planning module for collecting real-time user parking behavior data, real-time traffic status data, and parking lot occupancy rate data, and performing correlation analysis and path planning on the user parking behavior data, the real-time traffic status data, and the parking lot occupancy rate data to obtain candidate parking path data including a target parking lot and a navigation route; An analysis module is used to perform multi-dimensional cross-analysis on the user feedback data and the occupancy rate data collected in real time to obtain evaluation index data; The correction module is used to correct the candidate parking suggestion path data using the evaluation index data to obtain target parking suggestion path data.

9. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by a processor, the intelligent parking guidance method according to any one of claims 1 to 7 is implemented.

Images