Parking order analysis method and system based on big data

Abstract

The invention discloses a parking order analysis method and system based on big data, and relates to the technical field of parking order analysis, and the method specifically comprises the following steps: constructing a load conflict model for the path load index and path conflict risk coefficient of each generated departure path, generating the load conflict coefficient of each departure path, and calculating the load conflict coefficient of each departure path; analyzing after generation, and judging whether the load distribution of the departure path of the parking lot is balanced or not; when it is judged that the load distribution of the departure paths of the parking lot is unbalanced, a preset load conflict coefficient threshold interval is determined, and the preset load conflict coefficient threshold interval is compared with the load conflict coefficient of each departure path; and dividing each departure path into a high-conflict path, a normal path and a low-load path according to a comparison result. According to the invention, the problems of departure path dynamic analysis and optimal scheduling are solved, and the technical effects of path load balancing and efficient resource utilization are realized.

Description

Technical Field

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

[0002] Parking orders refer to records generated by users through the parking management system when vehicles are parked in parking lots, including parking time, parking space number, fee information, user vehicle information and other dynamic data. As a digital record of parking behavior, parking orders can reflect the use of parking resources, vehicle flow patterns and user demand characteristics. The significance of analyzing parking orders is that it can help identify the utilization efficiency of parking lots, user parking habits and parking space demand during peak hours, thereby providing data support for the management and optimization of parking lots. However, traditional analysis methods are limited by insufficient data volume and single analysis dimensions, making it difficult to dynamically capture real-time changing parking demands and potential patterns. Analyzing parking orders based on big data can integrate massive, multi-source real-time data, use data mining and prediction algorithms, deeply explore the complex patterns behind parking behaviors, accurately predict future parking demands, and dynamically optimize parking resource allocation, thereby achieving a comprehensive improvement in parking service efficiency and economic benefits.

[0003] The analysis process of existing parking order analysis technology based on big data mainly includes four aspects: data acquisition, dynamic monitoring, pattern recognition and behavior evaluation. First, use the parking management system or IoT devices to obtain multi-source data related to parking orders, such as order generation time, parking duration, user information and parking resource information, and combine external environmental factors to form a comprehensive dynamic data set. Then, through dynamic data monitoring, continue to pay attention to the changing trend of the number of parking orders, identify the peak and trough periods of parking demand, and observe its correlation with factors such as time and geographical location. On this basis, through in-depth analysis of order data, specific patterns of parking behavior are identified, such as high-frequency order areas and concentrated areas of long-term parking demand, so as to intuitively show the characteristics of parking resource usage. Finally, combined with the dynamic changes of parking orders, behavioral evaluation is carried out to analyze the operational performance of parking lots in different scenarios and the degree of satisfaction of user needs, providing an effective basis for the optimal allocation of parking resources. The core of this process is to reveal the characteristics of parking behavior and support refined management through dynamic monitoring and pattern recognition based on comprehensive data.

[0004] The prior art has the following deficiencies:

[0005] In intelligent parking lots, when vehicles leave the lot at the same time during peak hours, the parking lot needs to analyze the distribution of vehicle departure times in real time in order to optimize traffic scheduling. However, due to the complexity and variety of vehicle departure paths, each vehicle may choose a different path to leave, and the order data only records the end time of the order, and fails to effectively associate the vehicle's departure path information, resulting in the system being unable to accurately determine the concentration of departing vehicles on different paths and possible conflict points. This situation mainly occurs because the departure path label data has not been collected or incorporated into the existing analysis model. The existing technology can only make predictions based on simple time distribution statistics, and cannot be combined with the dynamic information of the departure path for comprehensive analysis. This will cause the parking lot to be unable to identify high-conflict paths or traffic concentration areas in a timely manner, which will lead to traffic overload on some paths, increase the risk of traffic congestion, and not only affect the rapid departure of vehicles, but may also cause spillover effects on traffic around the parking lot, further reducing the user's parking experience and the overall operation efficiency of the parking lot.

[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0007] The purpose of the present invention is to provide a parking order analysis method and system based on big data to solve the problems in the above-mentioned background technology.

[0008] In order to achieve the above object, the present invention provides the following technical solution: a parking order analysis method based on big data, specifically comprising the following steps:

[0009] According to the parking order information in the parking lot, including the order generation time, end time and parking space location, combined with the overall layout of the parking lot, the historical vehicle driving trajectory and exit distribution information, the parking lot's departure path is determined, and a unique path number and corresponding label are assigned to each departure path to build a parking lot's departure path network model;

[0010] Obtain the path dynamic information of each departure path in the parking lot in real time, analyze it after acquisition, and generate the path load index and path conflict risk coefficient of each departure path respectively;

[0011] A load conflict model is constructed based on the path load index and path conflict risk coefficient of each generated departure path, and the load conflict coefficient of each departure path is generated. After the generation, the load conflict coefficient is analyzed to determine whether the load distribution of the departure paths of the parking lot is balanced;

[0012] When it is determined that the load distribution of the departure paths of the parking lot is uneven, a preset load conflict coefficient threshold interval is determined, and the preset load conflict coefficient threshold interval is compared with the load conflict coefficient of each departure path, and each departure path is divided into a high-conflict path, a normal path, and a low-load path according to the comparison result;

[0013] According to the classification results of each departure path, corresponding optimization scheduling measures are taken for different types of paths;

[0014] Continuously update path dynamic information and load conflict models, dynamically adjust scheduling strategies based on real-time data, and ensure load distribution balance and resource utilization efficiency on departure paths.

[0015] Preferably, the path dynamic information of each departure path in the parking lot is obtained in real time, and analyzed after acquisition to generate the path load index and path conflict risk coefficient of each departure path respectively, which specifically includes the following steps:

[0016] Obtain the dynamic path information of each exit path in the parking lot in real time and perform preprocessing after acquisition;

[0017] Extracting load characteristic information and interactive dynamic information from the pre-processed path dynamic information of each departure path;

[0018] The load characteristic information and interactive dynamic information in the extracted path dynamic information of each departure path are analyzed to generate a path load index and a path conflict risk coefficient of each departure path respectively.

[0019] Preferably, the logic for obtaining the path load index of each departure path is as follows:

[0020] The load characteristic information of the preprocessed path dynamic information of each departure path is extracted, including the number of vehicles passing through each departure path at different times within a period of time, the maximum number of vehicles allowed to pass through each departure path per unit time in the historical records, and the total length of each departure path from the starting point to the exit, and they are marked as FLH i and PL i , represents the number of vehicles passing through the i-th departure path at time m within a period of time, FLH i In the historical records, PL represents the maximum number of vehicles allowed to pass through the i-th departure path per unit time. i represents the total length of the ith departure path from the starting point to the exit, i = 1, 2, 3, ..., k, m = 1, 2, 3, ..., g, k and g are both positive integers;

[0021] Calculate the path load index of each departure path. The specific calculation formula is as follows:

[0022]

[0023] Where PLI i is the path load index of the ith departure path.

[0024] Preferably, the logic for obtaining the path conflict risk coefficient of each departure path is as follows:

[0025] The interactive dynamic information in the preprocessed path dynamic information of each departure path is extracted, including the average vehicle flow at the intersection of each departure path with other paths at different times within a period of time, the average waiting time of vehicles at the intersection of each departure path with other paths, and the vehicle passing rate at the exit of each departure path, and they are calibrated as and represents the average vehicle flow at the intersection of the ith departure path and other paths at time n within a period of time, It represents the average waiting time of vehicles at the intersection of the i-th departure path and other paths at time n within a period of time. represents the vehicle passing rate at the exit of the i-th departure path at time n within a period of time, i = 1, 2, 3, ..., k, n = 1, 2, 3, ..., h, k and h are both positive integers;

[0026] Calculate the path conflict risk coefficient of each departure path. The specific calculation formula is as follows:

[0027]

[0028] Where, PCRC i is the path conflict risk coefficient of the i-th departure path.

[0029] Preferably, a load conflict model is constructed for the path load index and path conflict risk coefficient of each generated departure path, a load conflict coefficient of each departure path is generated, and an analysis is performed after the generation to determine whether the load distribution of the departure paths of the parking lot is balanced, specifically including the following steps:

[0030] The path load index, path conflict risk coefficient and corresponding load conflict coefficient of several departure paths generated in the past period of time are collected and calibrated as and x represents the path load index, path conflict risk coefficient and the number of the corresponding load conflict coefficient of each departure path generated in the past period of time, x=3, 4, 5, ..., d, d is a positive integer, and the data collected in the past period of time is formed into a historical data set;

[0031] Select the multivariate regression model as the load conflict model, and train it through the historical data set to determine the value of the regression coefficient according to the formula:

[0032]

[0033] In the formula, β 0 , β 1 and β 2 is the regression coefficient;

[0034] By minimizing the error between the predicted value and the actual value, the regression coefficient is optimized and the regression coefficient β is finally determined. 0 , β 1 and β 2 The value of

[0035] Using the final regression coefficient, the load conflict model is constructed and the path load index PLI of each departure path generated in real time is input. i and path conflict risk coefficient PCRC i , generate the load conflict coefficient LCC of each departure path in real time i ;

[0036] Calculate the standard deviation of the load conflict coefficient of each departure path. The specific calculation formula is as follows:

[0037]

[0038] In the formula, σ LCC is the standard deviation of the load conflict coefficient of each departure path;

[0039] Determine the standard deviation threshold of the load conflict coefficient for each pre-set departure path After determining, the standard deviation σ of the load conflict coefficient of each departure path LCC Compare and judge whether the load distribution of the parking lot's departure path is balanced according to the comparison results. The specific comparison and analysis is as follows:

[0040] like The parking lot’s departure paths have a balanced load distribution;

[0041] like The load distribution on the departure paths of the parking lot is uneven.

[0042] Preferably, when it is determined that the load distribution of the departure path of the parking lot is uneven, a preset load conflict coefficient threshold interval [LCC min , LCC max ] and compare the pre-set load conflict coefficient threshold interval with the load conflict coefficient LCC of each departure path iCompare and divide each departure path into high-conflict path, normal path and low-load path according to the comparison results. The specific comparison analysis and division are as follows:

[0043] If LCC i <LCC min , dividing the departure path into a low-load path;

[0044] If LCC min ≤LCC i ≤LCC max , divide the departure path into a normal path;

[0045] If LCC i >LCC max , classifying the departure path as a high-conflict path.

[0046] Preferably, according to the division results of each departure path, corresponding optimization scheduling measures are taken for different types of paths, specifically:

[0047] For the paths that are divided into high-conflict paths, the specific measures taken are: implement diversion and control strategies, dynamically adjust the guidance information of the path guidance screen, guide some vehicles to choose low-load paths, and optimize the timing of signal lights to increase the traffic efficiency of the path;

[0048] For the paths that are classified as normal paths, the specific measures taken are: maintain the existing scheduling strategy and guidance measures, regularly monitor the changes in its load conflict coefficient, ensure that the path load is in a stable state, and no additional intervention measures are required to ensure the smooth operation of the overall parking lot resources;

[0049] For paths that are divided into low-load paths, the specific measures taken are: give priority to guiding more vehicles into this path, achieve reasonable vehicle diversion through path guidance screens and dynamic traffic distribution, and at the same time reduce its idle resources and optimize the overall path utilization of the parking lot.

[0050] Preferably, the parking order analysis system based on big data includes a departure path network construction module, a path dynamic information analysis module, a load conflict model generation module, a path load classification module, a path optimization scheduling module and a dynamic update scheduling module;

[0051] The departure path network construction module determines the departure path of the parking lot based on the parking order information in the parking lot, including the order generation time, end time and parking space location, combined with the overall layout of the parking lot, the historical vehicle driving trajectory and the exit distribution information, and assigns a unique path number and corresponding label to each departure path to build the departure path network model of the parking lot;

[0052] The path dynamic information analysis module obtains the path dynamic information of each departure path in the parking lot in real time, analyzes it after acquisition, and generates the path load index and path conflict risk coefficient of each departure path respectively;

[0053] The load conflict model generation module constructs a load conflict model based on the path load index and path conflict risk coefficient of each generated departure path, generates the load conflict coefficient of each departure path, and analyzes it after generation to determine whether the load distribution of the departure paths of the parking lot is balanced;

[0054] The path load classification module determines a preset load conflict coefficient threshold interval when it is determined that the load distribution of the departure paths of the parking lot is uneven, and compares the preset load conflict coefficient threshold interval with the load conflict coefficient of each departure path, and divides each departure path into a high-conflict path, a normal path, and a low-load path according to the comparison result;

[0055] The path optimization and scheduling module takes corresponding optimization and scheduling measures for different categories of paths according to the classification results of each departure path;

[0056] Dynamically update the scheduling module, continuously update the path dynamic information and load conflict model, dynamically adjust the scheduling strategy according to real-time data, and ensure the load distribution balance and resource utilization efficiency of the departure path.

[0057] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0058] 1. The present invention effectively solves the problem in the prior art that it is impossible to associate vehicle order data with dynamic information of the departure path by constructing a network model of the parking lot's departure path and assigning a unique number and label to each path. This innovation enables the parking lot to accurately determine the vehicle's departure path based on order data and layout information, thereby providing a structured basis for the collection and analysis of dynamic path information. The advantage is that the establishment of the departure path network model makes the path management of the parking lot more refined and controllable, significantly improves the accuracy of data collection and analysis, and lays the foundation for subsequent model construction and scheduling optimization.

[0059] 2. The present invention obtains path dynamic information in real time and generates a path load index and a path conflict risk coefficient, and calculates the load conflict coefficient in combination with the load conflict model. The system can accurately evaluate the load status and conflict risk of each path. The application of this multivariate regression model can not only comprehensively consider multiple dynamic factors of path load and conflict, but also determine whether the path load distribution is balanced by dynamically calculating the load conflict coefficient. The significant effect of this technical solution is that it can timely discover high-conflict paths and low-load paths, providing a data-driven scientific basis for path optimization scheduling. Its advantages are: compared with traditional simple time distribution statistics, this method more comprehensively analyzes dynamic path characteristics, and has a high degree of real-time and accuracy, effectively reducing the occurrence of uneven path load distribution and congestion.

[0060] 3. The present invention realizes efficient utilization of path resources and load balancing through the implementation of path classification and dynamic scheduling optimization strategies. The optimized scheduling measures taken for high-conflict paths, normal paths and low-load paths respectively enable the parking lot to dynamically adjust the traffic distribution during peak hours, effectively alleviate the congestion pressure of high-conflict paths, and improve the utilization rate of low-load paths. Combined with the function of continuously updating path dynamic information and load conflict models, the solution can achieve real-time optimization and dynamic adaptation, thereby ensuring that the overall operating efficiency and user experience of the parking lot are significantly improved. The advantage of this design is that it not only improves the utilization rate of path resources, but also improves the scheduling flexibility and intelligence level of the parking lot, providing users with a smoother and more efficient departure experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0062] Figure 1 It is a flow chart of the parking order analysis method and system based on big data of the present invention.

[0063] Figure 2 This is a module schematic diagram of the parking order analysis method and system based on big data of the present invention. DETAILED DESCRIPTION

[0064] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0065] The present invention provides Figure 1 The parking order analysis method based on big data shown in the figure specifically includes the following steps:

[0066] According to the parking order information in the parking lot, including the order generation time, end time and parking space location, combined with the overall layout of the parking lot, the historical vehicle driving trajectory and exit distribution information, the parking lot's departure path is determined, and a unique path number and corresponding label are assigned to each departure path to build a parking lot's departure path network model;

[0067] To determine the exit path of a parking lot based on parking order information and other relevant data, the following methods can be used: First, extract the parking starting point information of each vehicle through parking order data (including order generation time, end time and parking space location) to determine the vehicle's departure time and location. Combined with the overall layout information of the parking lot, obtain the connection relationship and exit distribution of the lanes in the parking lot. Then, through the historical driving trajectory data (recorded by sensors, cameras or license plate recognition systems), analyze the actual path selection pattern of the vehicle from the parking position to the exit and match it with the structure of the parking lot. Using this data, predict the optimal path for the vehicle to leave through a path recognition algorithm (such as a shortest path or weighted path algorithm). Finally, determine the departure path of each vehicle in the parking lot through model analysis and path prediction.

[0068] Assigning a unique path number and corresponding label to each departure path and building a network model of the departure path of the parking lot can be achieved in the following ways: numbering the determined departure path set according to the exit number, path flow or historical usage frequency, for example, based on the exit number + path sequence number (such as "E1-P1" represents path 1 of exit 1). At the same time, according to the characteristics of the path (such as path flow, length or historical usage rate), assign a label to each path (such as "main path", "low load path", etc.). Then, based on the parking lot layout diagram, the path is represented as a directed graph, in which the node represents the intersection or exit, the edge represents the path, and the edge weight reflects the characteristics of the path (such as travel time or flow). The network model is dynamically updated through the algorithm so that it can reflect the real-time status of the path and provide support for subsequent path classification and optimization scheduling.

[0069] This design is to accurately determine the vehicle departure path and classify and manage it through parking order information, path numbering and network model, and solve the problem of uneven vehicle distribution and path conflict during the peak departure period of the parking lot. Through path labeling and modeling, it can support the dynamic calculation of path load and conflict risk, identify high-conflict paths and take corresponding optimization measures. The path network model provides the parking lot with the ability of global path management, ensuring that vehicles can be evenly distributed on each departure path, improving the efficiency of traffic scheduling during peak hours, reducing congestion, and improving user experience.

[0070] Obtain the path dynamic information of each departure path in the parking lot in real time, analyze it after acquisition, and generate the path load index and path conflict risk coefficient of each departure path respectively;

[0071] In this embodiment, the path dynamic information of each departure path in the parking lot is obtained in real time, and analyzed after acquisition to generate the path load index and path conflict risk coefficient of each departure path respectively, which specifically includes the following steps:

[0072] Obtain the dynamic path information of each exit path in the parking lot in real time and perform preprocessing after acquisition;

[0073] Real-time acquisition of path dynamic information of each exit path in the parking lot can be achieved through the sensor network, camera system and license plate recognition equipment deployed in the parking lot. These hardware devices collect data such as traffic flow, vehicle movement trajectory, intersection status and exit traffic rate of each exit path in real time. For example, sensors can record the number and speed of passing vehicles on each path, cameras can capture the waiting time and path selection behavior of vehicles at intersections, and license plate recognition devices record the complete trajectory of vehicles from entering the parking space to leaving the path. These data are transmitted to the data analysis system through the wireless network, summarized into path dynamic information, and the software system processes the received data in real time to ensure the continuity and integrity of the information.

[0074] Preprocessing is performed to improve the accuracy and reliability of data analysis, while reducing the interference of redundant and erroneous information on subsequent calculations. Specific preprocessing steps include: data denoising, format conversion and standardization. Data denoising can remove abnormal sensor signals or camera recognition errors through filtering algorithms; format conversion is to convert data collected by different devices (such as traffic, trajectory, waiting time) into a calculable numerical form; standardization uses normalization methods to adjust the numerical range of different data to a unified scale for comprehensive analysis. For example, traffic data can be smoothed through a time window, trajectory data can be converted to path numbers, and waiting time data can be normalized by seconds. The preprocessing process is completed through the software's automated analysis module to ensure that an accurate data basis is provided for subsequent analysis efficiently and in real time.

[0075] Extracting load characteristic information and interactive dynamic information from the pre-processed path dynamic information of each departure path;

[0076] The load characteristic information and interactive dynamic information in the extracted path dynamic information of each departure path are analyzed to generate a path load index and a path conflict risk coefficient of each departure path respectively.

[0077] In this embodiment, the logic for obtaining the path load index of each departure path is as follows:

[0078] The load characteristic information of the preprocessed path dynamic information of each departure path is extracted, including the number of vehicles passing through each departure path at different times within a period of time, the maximum number of vehicles allowed to pass through each departure path per unit time in the historical records, and the total length of each departure path from the starting point to the exit, and they are marked as FLH i and PL i , represents the number of vehicles passing through the i-th departure path at time m within a period of time, FLH i In the historical records, PL represents the maximum number of vehicles allowed to pass through the i-th departure path per unit time. i represents the total length of the ith departure path from the starting point to the exit, i = 1, 2, 3, ..., k, m = 1, 2, 3, ..., g, k and g are both positive integers;

[0079] Extracting load characteristic information from the path dynamic information of each preprocessed departure path can be achieved by path data classification and labeling. First, the path dynamic information is denoised and standardized through the preprocessing module, and data from different sources (such as sensor flow records, camera trajectory capture, etc.) are classified and stored in a unified data warehouse. Then, all data are associated with each departure path through the departure path number, for example, the number of vehicles passing, path length and other information are classified and sorted according to the path number to form a corresponding load characteristic information set. Finally, using the data extraction algorithm, by setting the time window, the real-time data and historical data within a specific time period, as well as the static geometric characteristics of the path, are filtered out to generate complete load characteristic information including the real-time number of vehicles, the historical maximum passing capacity and the path length.

[0080] Three types of quantitative data, namely, the number of vehicles passing through each departure path at different times within a period of time, the maximum number of vehicles allowed to pass through each departure path per unit time in historical records, and the total length of each departure path from the starting point to the exit, can be obtained from different data sources. The first type of data (the number of vehicles at different times) is collected in real time by sensors or cameras deployed on the path, for example, the passing time of each vehicle is recorded at the entrance and exit of the path, and the statistics are segmented by time intervals; the second type of data (the historical maximum passing capacity) can be obtained by analyzing the peak traffic records of the same path in the historical database, for example, the highest traffic value per unit time (such as one minute) is extracted from long-term data as a historical record; the third type of data (the total length of the path) can be measured by the digital map of the parking lot or the path modeling algorithm. The straight-line distance between the starting point and the exit of the path is measured and accumulated according to the actual curvature of the path. These data are automatically analyzed and matched to the corresponding path number by software to ensure that the generated information is complete and accurate.

[0081] Calculate the path load index of each departure path. The specific calculation formula is as follows:

[0082]

[0083] Where PLI i is the path load index of the ith departure path.

[0084] The calculation logic of this formula comprehensively considers the comprehensive influence of the real-time traffic characteristics, historical carrying capacity, path length and time distribution of the path, and reflects its rationality in steps. First, The impact of high traffic on the load index is magnified, highlighting the pressure of high load on the path, and combining the historical maximum traffic FLH i and path length PL i As the denominator, the path carrying capacity and spatial characteristics are normalized to balance the differences in static and dynamic characteristics of the path. Then, by accumulating and averaging the data of multiple time periods, the dynamic load changes of the path in different time periods are captured to ensure that the index can reflect the average state of the entire time period. Finally, the square root is taken to adjust the index range to make the result smoother and easier to interpret. This calculation method not only focuses on the impact of instantaneous high load on the path, but also reflects the dynamic changes of the entire time period, and scientifically and comprehensively evaluates the load characteristics of the path.

[0085] Path load index PLI of the i-th departure path iThe size of directly reflects the load pressure of the path in a specific time period, and is one of the key bases for judging whether the load distribution of the parking lot's departure path is balanced. If there is a large difference in the path load index among all paths (such as the load index of some paths is significantly higher than that of other paths), it means that the load distribution of the parking lot's departure path is uneven, and some paths may be overloaded or waste resources. By comparing the load index of each path with the set evaluation threshold interval, the paths can be further divided into high-conflict paths (the load index exceeds the upper limit, indicating that the path is close to an overload state), normal paths (the load index is within a reasonable range, and the path state is stable) and low-load paths (the load index is lower than the lower limit, indicating that the path is not fully utilized). This classification provides an accurate basis for subsequent optimization and scheduling, such as reducing the pressure of high-conflict paths through dynamic diversion, giving priority to guiding vehicles to low-load paths, and improving overall resource utilization efficiency.

[0086] In this embodiment, the logic for obtaining the path conflict risk coefficient of each departure path is as follows:

[0087] The interactive dynamic information in the preprocessed path dynamic information of each departure path is extracted, including the average vehicle flow at the intersection of each departure path with other paths at different times within a period of time, the average waiting time of vehicles at the intersection of each departure path with other paths, and the vehicle passing rate at the exit of each departure path, and they are calibrated as and represents the average vehicle flow at the intersection of the ith departure path and other paths at time n within a period of time, It represents the average waiting time of vehicles at the intersection of the i-th departure path and other paths at time n within a period of time. represents the vehicle passing rate at the exit of the i-th departure path at time n within a period of time, i = 1, 2, 3, ..., k, n = 1, 2, 3, ..., h, k and h are both positive integers;

[0088] Extracting interactive dynamic information from the path dynamic information of each preprocessed departure path can be achieved through multi-source data fusion and dynamic analysis methods of path intersections. First, through sensors, cameras and license plate recognition devices deployed at path intersections and exits, basic data such as vehicle flow, vehicle waiting time and traffic rate are collected in real time. These data are classified and matched by software and associated with specific path numbers and intersection numbers. Then, using time segmentation analysis technology, time-sharing statistics are performed on data in different time periods, and the average values ​​of intersection flow, waiting time and exit traffic rate of each path are calculated respectively. These calculation results are written into the data warehouse in real time through a dynamic update mechanism, and the load interaction characteristic data is extracted through a data extraction algorithm to form a complete interactive dynamic information set.

[0089] The three types of quantitative data, namely, the average vehicle flow at the intersection of each departure path with other paths at different times over a period of time, the average waiting time of vehicles at the intersection of each departure path with other paths, and the vehicle throughput rate at the exit of each departure path, can be obtained in the following ways: the first type of data (average vehicle flow at the intersection of paths) records the number of vehicles per unit time through sensors or cameras at the intersection of paths, and averages the data of different time periods according to the set time window; the second type of data (average waiting time of vehicles) captures the actual waiting time of each vehicle at the intersection through cameras, and then calculates the average within the time period through software; the third type of data (vehicle throughput rate at the exit) records the number of vehicles passing through the exit and the corresponding time interval through sensors at the exit, and the software system calculates the throughput rate per unit time and stores it. These data are updated in real time and dynamically associated with the specific intersection and exit characteristics of each path through data fusion and automated analysis processes.

[0090] Calculate the path conflict risk coefficient of each departure path. The specific calculation formula is as follows:

[0091]

[0092] Where, PCRC i is the path conflict risk coefficient of the i-th departure path.

[0093] The formula scientifically evaluates the conflict risk of the path by comprehensively considering the vehicle flow at the intersection, the square of the waiting time, and the exit traffic rate. Average vehicle flow at the intersection It reflects the density of path interaction and is the basis for measuring conflict risk. This amplifies the effect of long wait times on conflict risk, as longer wait times are often directly associated with higher conflict likelihood; exit traffic rate As the denominator, the diversion capacity of the exit is measured. The higher the traffic rate, the stronger the ability to alleviate conflicts and the lower the risk. In addition, the conflict risks of multiple time periods are smoothed by the time period average to ensure that the formula can comprehensively reflect the dynamic characteristics of the path in the entire time period. This design is both dynamic and comprehensive, which helps to accurately assess the conflict risk of the path and provide data support for subsequent path optimization.

[0094] The path conflict risk coefficient PRC of the i-th departure path iThe size of directly reflects the dynamic risk level of the path at the intersection, and is closely related to the judgment of the balance of the overall path load distribution of the parking lot. When there are large differences in the conflict risk coefficients of each path, it means that some paths may face high risks (such as dense traffic at the intersection or long waiting time), while other paths are less burdened, reflecting the uneven load distribution between paths. By comparing the conflict risk coefficient of each path with the set evaluation threshold interval, the paths can be further classified into high-conflict paths (risk coefficient is higher than the upper threshold limit, and priority regulation is required to relieve pressure), normal paths (risk coefficient is within a reasonable range, no special intervention is required), and low-load paths (risk coefficient is lower than the lower threshold limit, and resources may not be fully utilized). This classification provides an accurate basis for subsequent path optimization and traffic scheduling, helping to balance the overall departure load distribution of the parking lot, improve traffic efficiency and reduce conflict risks.

[0095] A load conflict model is constructed based on the path load index and path conflict risk coefficient of each generated departure path, and the load conflict coefficient of each departure path is generated. After the generation, the load conflict coefficient is analyzed to determine whether the load distribution of the departure paths of the parking lot is balanced;

[0096] In this embodiment, a load conflict model is constructed based on the path load index and path conflict risk coefficient of each generated departure path, and the load conflict coefficient of each departure path is generated. After the generation, an analysis is performed to determine whether the load distribution of the departure paths of the parking lot is balanced. Specifically, the following steps are included:

[0097] The path load index, path conflict risk coefficient and corresponding load conflict coefficient of several departure paths generated in the past period of time are collected and calibrated as and x represents the path load index, path conflict risk coefficient and the number of the corresponding load conflict coefficient of each departure path generated in the past period of time, x=3, 4, 5, ..., d, d is a positive integer, and the data collected in the past period of time is formed into a historical data set;

[0098] The collection of path load index, path conflict risk coefficient and corresponding load conflict coefficient of each departure path generated in the past period of time can be achieved through the data storage and scheduling management system. First, the path data analysis system of the parking lot will calculate the path load index and path conflict risk coefficient of each path in real time, and write these results into the database immediately after generation. At the same time, the corresponding load conflict coefficient is calculated through the load conflict model and stored in the database. The system sets the time window for data storage (for example, the last week or month), classifies and archives all data according to timestamps and path numbers, and ensures that the data of each path in different time periods is complete and traceable. During the collection process, the archived path data set in the database is called through the scheduling module of the software, and the required historical data set is filtered out according to the set time range and path number conditions. The data is packaged into the analysis module through the batch export function for subsequent model training and verification to ensure the real-time and accuracy of data collection.

[0099] The value range of x is limited to positive integers greater than or equal to 3 because three regression coefficients β need to be calculated in the target equation. 0 , β 1 and β 2 . In order to accurately determine the values ​​of these three regression coefficients, at least three independent samples (i.e., x ≥ 3) are required to form three independent equations to ensure that the system of equations has a unique solution. If the number of samples is less than 3, the degree of freedom of the system of equations is insufficient and the unique regression coefficient cannot be solved. In addition, selecting x ≥ 3 can also ensure that the model has basic computing capabilities under minimum conditions, while providing basic fitting capabilities and error optimization space for the regression model. Further increasing the value of x (i.e., more samples) can improve the stability and fitting accuracy of the model, but at least three samples are the minimum requirement for determining the regression coefficients, so this limitation is reasonable and necessary.

[0100] Select the multivariate regression model as the load conflict model, and train it through the historical data set to determine the value of the regression coefficient according to the formula:

[0101]

[0102] In the formula, β 0 , β 1 and β 2 is the regression coefficient;

[0103] The multivariate regression model is a statistical analysis method used to analyze the impact of multiple independent variables (such as path load index and path conflict risk coefficient) on the dependent variable (load conflict coefficient), and to express the relationship between the independent variables and the dependent variable by establishing a linear equation. The reason for choosing the multivariate regression model as the load conflict model is that this method can comprehensively consider the relationship between multiple influencing factors and optimize the regression coefficient through historical data training, so that the model's prediction results are closer to the actual situation. Through the training of historical data sets, the regression model can minimize the error between the predicted value and the actual value, and determine the regression coefficient β 0 , β 1 and β 2 The optimal value of . 0 is the intercept of the model, which indicates the basic value of the load conflict coefficient when both the path load index and the path conflict risk coefficient are zero; β 1 is the regression coefficient of the path load index, which measures the linear effect of the path load index on the load conflict coefficient; β 2 is the regression coefficient of the path conflict risk coefficient, which indicates the influence of the path conflict risk coefficient on the load conflict coefficient. In this way, a scientific and accurate load conflict model can be constructed, which provides an important reference for the analysis of path status and scheduling optimization.

[0104] By minimizing the error between the predicted value and the actual value, the regression coefficient is optimized and the regression coefficient β is finally determined. 0 , β 1 and β 2 The value of

[0105] The specific implementation method of optimizing the regression coefficient by minimizing the error between the predicted value and the actual value is based on the error optimization algorithm of multivariate linear regression (such as the least squares method). First, the path load index and path conflict risk coefficient in the historical data set are used as independent variables, and the corresponding load conflict coefficient in the historical data is used as the dependent variable and substituted into the initial regression model. Then, the error between the model predicted value and the actual value is calculated (usually expressed as a square error). Then, the regression coefficient β is adjusted through iterative optimization. 0 , β 1 and β 2 The error value is gradually reduced until the error converges to the minimum value, and the optimal regression coefficient is determined. The reason for this is to ensure that the model can reflect the actual situation to the greatest extent, making the linear influence relationship between the path load index and the path conflict risk coefficient on the load conflict coefficient more accurate. By minimizing the error, the model has a higher fit to the historical data, thereby improving its prediction accuracy in the actual path state evaluation, and providing a scientific basis for subsequent path state analysis and optimized scheduling.

[0106] Using the final regression coefficient, the load conflict model is constructed and the path load index PLI of each departure path generated in real time is input. i and path conflict risk coefficient PCRC i , generate the load conflict coefficient LCC of each departure path in real time i ;

[0107] Calculate the standard deviation of the load conflict coefficient of each departure path. The specific calculation formula is as follows:

[0108]

[0109] In the formula, σ LCC is the standard deviation of the load conflict coefficient of each departure path;

[0110] Determine the standard deviation threshold of the load conflict coefficient for each pre-set departure path After determining, the standard deviation σ of the load conflict coefficient of each departure path LCC Compare and judge whether the load distribution of the parking lot's departure path is balanced according to the comparison results. The specific comparison and analysis is as follows:

[0111] like The parking lot’s departure paths have a balanced load distribution;

[0112] When the standard deviation of the load conflict coefficients of each departure path is less than or equal to the preset standard deviation threshold, it means that the load distribution of the parking lot departure paths is balanced and the load conflict coefficients of each path are relatively small. This shows that the vehicle departure flow is reasonably distributed among the paths, without obvious overload or waste of resources. The impact is that the parking lot exit congestion risk is low, the vehicle traffic efficiency is high, and no additional adjustments are required to the path scheduling. The existing departure path management strategy can be maintained, thereby saving resources and improving operational efficiency.

[0113] like The load distribution on the departure paths of the parking lot is uneven.

[0114] When the standard deviation of the load conflict coefficients of each departure path is greater than the preset standard deviation threshold, it means that the load distribution of the parking lot departure paths is uneven, some paths may have high conflicts or high loads, while other paths may have insufficient resource utilization. This will increase the risk of congestion on some paths, affecting the rapid departure of vehicles, and may have a negative impact on the overall traffic efficiency and user experience of the parking lot. To solve this problem, it is necessary to divert high-load paths through dynamic scheduling strategies and direct part of the traffic to low-load paths, thereby rebalancing the load distribution of the paths.

[0115] Determining the standard deviation threshold of the load conflict coefficient of each departure path can be achieved through statistical analysis of historical data and dynamic adjustment. First, the software system is used to extract the load conflict coefficients of all paths in each time period from the historical operation data of the parking lot, calculate the standard deviation of the load conflict coefficient in each time period, and generate a set of historical standard deviation data sets. Then, through statistical analysis methods (such as confidence interval analysis based on the normal distribution assumption), the average value and fluctuation range of the standard deviation in the historical data are determined, and the average value plus an appropriate confidence interval (such as a 95% confidence interval) is used as the initial standard deviation threshold. On this basis, the system can introduce a dynamic adjustment mechanism to dynamically optimize the threshold by real-time monitoring the standard deviation of the load conflict coefficient of the current path and combining the trend of changes in historical standard deviation data. For example, when it is found that the current standard deviation deviates from the historical mean for a long time, the software system can retrain and update the threshold through a machine learning model to ensure its dynamic adaptability to the operation status of the parking lot, thereby accurately guiding the judgment of load distribution balance.

[0116] When it is determined that the load distribution of the departure paths of the parking lot is uneven, a preset load conflict coefficient threshold interval is determined, and the preset load conflict coefficient threshold interval is compared with the load conflict coefficient of each departure path, and each departure path is divided into a high-conflict path, a normal path, and a low-load path according to the comparison result;

[0117] In this embodiment, when it is determined that the load distribution of the departure path of the parking lot is uneven, a preset load conflict coefficient threshold interval [LCC min , LCC max ] and compare the pre-set load conflict coefficient threshold interval with the load conflict coefficient LCC of each departure path i Compare and divide each departure path into high-conflict path, normal path and low-load path according to the comparison results. The specific comparison analysis and division are as follows:

[0118] If LCC i <LCC min , dividing the departure path into a low-load path;

[0119] The classification criteria for low-load paths are based on the load conflict coefficient of the path being lower than the preset minimum threshold, indicating that the actual load pressure and conflict risk of the path are low. This classification is to identify paths with insufficient resource utilization and avoid idle paths causing a decrease in the overall resource allocation efficiency of the parking lot. By clearly classifying low-load paths, more departing vehicles can be guided to enter these paths in the scheduling strategy, thereby making full use of existing resources and improving the overall departure efficiency of the parking lot.

[0120] If LCC min ≤LCCi ≤LCC max , divide the departure path into a normal path;

[0121] The normal path is divided into two categories: the path load conflict coefficient is within the preset threshold range, indicating that the load and conflict status of the path are within an acceptable range, with neither obvious overload nor idle resources. This division is to ensure the stability of the path operation and maintain the normal traffic order of the parking lot. For normal paths, the scheduling strategy can maintain the existing guidance and allocation mode, avoid unnecessary adjustments, and ensure that the parking lot resources operate efficiently in a stable state.

[0122] If LCC i >LCC max , classifying the departure path as a high-conflict path.

[0123] The classification standard for high-conflict paths is that the path load conflict coefficient exceeds the preset maximum threshold, indicating that the load pressure or conflict risk of the path is too high, which may cause congestion or delays when vehicles leave the parking lot. This classification is to quickly identify high-risk paths and avoid further deterioration of congestion on these paths. By clearly classifying high-conflict paths, diversion and control measures can be taken to reduce their load pressure and optimize path distribution, thereby improving the overall traffic efficiency and user experience of the parking lot.

[0124] The preset load conflict coefficient threshold interval [LCC min , LCC max ] can be determined by combining historical data statistical analysis and real-time dynamic adjustment. First, the software system extracts the load conflict coefficient of each path from the historical operation data of the parking lot, calculates the maximum, minimum, mean and standard deviation of each time period, and generates a complete historical distribution data set. Then, using statistical analysis methods (such as percentile-based interval division or confidence interval calculation based on normal distribution assumptions), set an appropriate interval, such as [LCC min , LCC max ] are defined as the 5% and 95% percentiles in the historical distribution, respectively, to ensure that the conflict coefficients of most paths fall within this interval. In order to enhance dynamic adaptability, the software system can monitor the load conflict coefficient distribution of the current path in real time and compare it with the historical distribution. If a long-term deviation trend is found (such as a significant increase or decrease in path pressure), the upper and lower limits of the interval are automatically adjusted through a dynamic update algorithm to more accurately reflect the current parking lot operation status and guide path classification and scheduling optimization.

[0125] According to the classification results of each departure path, corresponding optimization scheduling measures are taken for different types of paths;

[0126] In this embodiment, according to the classification results of each departure path, corresponding optimization scheduling measures are taken for different types of paths, specifically:

[0127] For the paths that are classified as high-conflict paths, the specific measures taken are: implement diversion and control strategies, dynamically adjust the guidance information of the path guidance screen, guide some vehicles to choose low-load paths, optimize the timing of traffic lights, increase the traffic efficiency of the path, and restrict some vehicles from entering the path when necessary to reduce its load pressure and conflict risks;

[0128] The optimization of high-conflict paths can be achieved by implementing diversion control strategies, including dynamic path guidance screens, signal light optimization, and flow restrictions. First, the dynamic path guidance system of the parking lot obtains the load conflict coefficient of each path in real time and marks the high-conflict path as the path that needs to reduce traffic. The software system dynamically adjusts the display information of the path guidance screen according to real-time monitoring data, such as giving priority to recommending low-load paths as exits to divert some vehicles. Secondly, through the optimization of signal light timing, the green light time of other low-conflict paths is extended, and the signal passage time of high-conflict paths is shortened to reduce their load pressure. Finally, when necessary, traffic restriction measures are implemented. The software sets the quota of vehicles entering the path through the license plate recognition system to avoid further aggravation of congestion. The purpose of this is to reduce the load pressure of the path by comprehensively regulating the flow and traffic efficiency of the high-conflict path, thereby preventing serious congestion and optimizing the overall traffic distribution.

[0129] For the paths that are classified as normal paths, the specific measures taken are: maintain the existing scheduling strategy and guidance measures, regularly monitor the changes in its load conflict coefficient, ensure that the path load is in a stable state, and no additional intervention measures are required to ensure the smooth operation of the overall parking lot resources;

[0130] The key to optimizing the normal path is to maintain the status quo and conduct regular monitoring. Through the path monitoring module of the software, the load conflict coefficient data of the path is collected in real time and compared with historical data to ensure that the load status remains within a reasonable range. If the conflict coefficient is detected to be close to the upper and lower thresholds, an early warning is triggered to prompt the management staff to make further observations. This "maintenance + monitoring" strategy is implemented by the software automation system without the need for additional human intervention or resource investment. The reason for this is that the normal path is already in the best operating state, and intervention may cause waste of resources or unnecessary fluctuations. Maintaining the existing scheduling can ensure the smooth operation of the overall resources of the parking lot.

[0131] For paths that are divided into low-load paths, the specific measures taken are: give priority to guiding more vehicles to enter the path, achieve reasonable vehicle diversion through path guidance screens and dynamic traffic distribution, while reducing idle resources, optimizing the overall path utilization of the parking lot, and improving departure efficiency and resource allocation balance.

[0132] The optimization of low-load paths can be achieved through dynamic path guidance and traffic distribution algorithms. First, the software recommends low-load paths as exits through the dynamic path guidance screen, such as setting the low-load path as the preference of the navigation system or guidance screen to attract more vehicles to choose this path. Secondly, the traffic distribution algorithm is used to dynamically adjust the distribution of vehicles according to real-time traffic data to guide vehicles into low-load paths. The software can also gradually improve the utilization rate of low-load paths through learning and optimization, such as adjusting the weight of path recommendations. The purpose of this is to make full use of parking resources, reduce the idleness of low-load paths, and divert traffic from high-conflict paths, improve overall departure efficiency, and optimize the balance of resource allocation.

[0133] Continuously update path dynamic information and load conflict models, dynamically adjust scheduling strategies based on real-time data, and ensure load distribution balance and resource utilization efficiency on departure paths.

[0134] Continuously updating the path dynamic information and load conflict model can be achieved through real-time data collection, dynamic model update and intelligent scheduling strategy adjustment modules. First, the sensors, cameras and license plate recognition systems deployed in the parking lot collect dynamic data of each path in real time, including path flow, intersection status and exit traffic rate, and transmit them to the central database through the wireless network. Based on these data, the software system regularly updates the path load index and path conflict risk coefficient, and recalculates the load conflict coefficient using the latest data. Subsequently, the software updates the load conflict model according to these dynamic parameters, and optimizes the regression coefficient to adapt to the current path status by training a new regression model. Finally, combined with the latest model results, the path classification and scheduling strategy are dynamically adjusted. The entire process is completed by an automated analysis and control module to ensure real-time data updates and dynamic model optimization.

[0135] The continuous updating of path dynamic information and load conflict models is to adapt to the real-time changes in the load status of parking lot paths and ensure that the scheduling strategy can effectively respond to the fluctuations in vehicle flow and the dynamic changes in path conflicts. The load distribution of the parking lot's departure paths is usually affected by peak hours, concentrated vehicle departures, and emergencies (such as exit congestion). If the model and scheduling strategy cannot be updated in time, it may lead to path classification failure, increased congestion on high-conflict paths, and idleness of low-load paths. By continuously updating the model and data, real-time optimization of the scheduling strategy can be achieved, and the flow distribution of the path can be dynamically balanced, thereby effectively preventing resource waste and congestion risks and improving overall departure efficiency.

[0136] Continuous updates in this way ensure that the scheduling strategy is always based on the latest path status and model analysis results, providing scientific support for path flow allocation. Dynamically updated scheduling strategies can better respond to changes in vehicle flow and optimize the utilization of path resources. For example, the load pressure of high-conflict paths can be diverted in real time, and the utilization of low-load paths can be dynamically improved, so that the overall departure efficiency and resource allocation balance of the parking lot can be optimized. In addition, this continuous update and adjustment is also highly flexible and adaptable, which can provide a long-term stable and efficient solution for parking lot management, while improving the user's departure experience.

[0137] like Figure 2 The parking order analysis system based on big data shown includes a departure path network construction module, a path dynamic information analysis module, a load conflict model generation module, a path load classification module, a path optimization scheduling module and a dynamic update scheduling module;

[0138] The departure path network construction module determines the departure path of the parking lot based on the parking order information in the parking lot, including the order generation time, end time and parking space location, combined with the overall layout of the parking lot, the historical vehicle driving trajectory and the exit distribution information, and assigns a unique path number and corresponding label to each departure path to build the departure path network model of the parking lot;

[0139] The path dynamic information analysis module obtains the path dynamic information of each departure path in the parking lot in real time, analyzes it after acquisition, and generates the path load index and path conflict risk coefficient of each departure path respectively;

[0140] The load conflict model generation module constructs a load conflict model based on the path load index and path conflict risk coefficient of each generated departure path, generates the load conflict coefficient of each departure path, and analyzes it after generation to determine whether the load distribution of the departure paths of the parking lot is balanced;

[0141] The path load classification module determines a preset load conflict coefficient threshold interval when it is determined that the load distribution of the departure paths of the parking lot is uneven, and compares the preset load conflict coefficient threshold interval with the load conflict coefficient of each departure path, and divides each departure path into a high-conflict path, a normal path, and a low-load path according to the comparison result;

[0142] The path optimization and scheduling module takes corresponding optimization and scheduling measures for different categories of paths according to the classification results of each departure path;

[0143] Dynamically update the scheduling module, continuously update the path dynamic information and load conflict model, dynamically adjust the scheduling strategy according to real-time data, and ensure the load distribution balance and resource utilization efficiency of the departure path.

[0144] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0145] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center containing one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0146] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0147] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0148] In the several embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0149] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0150] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0151] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A parking order analysis method based on big data, characterized in that: The specific steps include: According to the parking order information in the parking lot, including the order generation time, end time and parking space location, combined with the overall layout of the parking lot, the historical vehicle driving trajectory and exit distribution information, the parking lot's departure path is determined, and a unique path number and corresponding label are assigned to each departure path to build a parking lot's departure path network model; Obtain the path dynamic information of each departure path in the parking lot in real time, analyze it after acquisition, and generate the path load index and path conflict risk coefficient of each departure path respectively; A load conflict model is constructed based on the path load index and path conflict risk coefficient of each generated departure path, and the load conflict coefficient of each departure path is generated. After the generation, the load conflict coefficient is analyzed to determine whether the load distribution of the departure paths of the parking lot is balanced; When it is determined that the load distribution of the departure paths of the parking lot is uneven, a preset load conflict coefficient threshold interval is determined, and the preset load conflict coefficient threshold interval is compared with the load conflict coefficient of each departure path, and each departure path is divided into a high-conflict path, a normal path, and a low-load path according to the comparison result; According to the classification results of each departure path, corresponding optimization scheduling measures are taken for different types of paths; Continuously update path dynamic information and load conflict models, dynamically adjust scheduling strategies based on real-time data, and ensure load distribution balance and resource utilization efficiency on departure paths.

2. The parking order analysis method based on big data according to claim 1 is characterized in that: The path dynamic information of each departure path in the parking lot is obtained in real time, and analyzed after acquisition to generate the path load index and path conflict risk coefficient of each departure path respectively, which specifically includes the following steps: Obtain the dynamic path information of each exit path in the parking lot in real time and perform preprocessing after acquisition; Extracting load characteristic information and interactive dynamic information from the pre-processed path dynamic information of each departure path; The load characteristic information and interactive dynamic information in the extracted path dynamic information of each departure path are analyzed to generate a path load index and a path conflict risk coefficient of each departure path respectively.

3. The parking order analysis method based on big data according to claim 2 is characterized in that: The logic for obtaining the path load index of each departure path is as follows: The load characteristic information of the preprocessed path dynamic information of each departure path is extracted, including the number of vehicles passing through each departure path at different times within a period of time, the maximum number of vehicles allowed to pass through each departure path per unit time in the historical records, and the total length of each departure path from the starting point to the exit, and they are marked as FLH i and PL i , represents the number of vehicles passing through the i-th departure path at time m within a period of time, FLH i In the historical records, PL represents the maximum number of vehicles allowed to pass through the i-th departure path per unit time. i represents the total length of the ith departure path from the starting point to the exit, i = 1, 2, 3, ..., k, m = 1, 2, 3, ..., g, k and g are both positive integers; Calculate the path load index of each departure path. The specific calculation formula is as follows: Where PLI i is the path load index of the i-th departure path.

4. The parking order analysis method based on big data according to claim 3 is characterized in that: The logic for obtaining the path conflict risk coefficient of each departure path is as follows: The interactive dynamic information in the preprocessed path dynamic information of each departure path is extracted, including the average vehicle flow at the intersection of each departure path with other paths at different times within a period of time, the average waiting time of vehicles at the intersection of each departure path with other paths, and the vehicle passing rate at the exit of each departure path, and they are calibrated as and represents the average vehicle flow at the intersection of the ith departure path and other paths at time n within a period of time, It represents the average waiting time of vehicles at the intersection of the i-th departure path and other paths at time n within a period of time. represents the vehicle passing rate at the exit of the ith departure path at time n within a period of time, 1, 2, 3, ..., k, n = 1, 2, 3, ..., h, k and h are both positive integers; Calculate the path conflict risk coefficient of each departure path. The specific calculation formula is as follows: Where, PCRC i is the path conflict risk coefficient of the i-th departure path.

5. The parking order analysis method based on big data according to claim 4 is characterized in that: A load conflict model is constructed based on the path load index and path conflict risk coefficient of each generated departure path, and the load conflict coefficient of each departure path is generated. After the generation, an analysis is performed to determine whether the load distribution of the departure paths of the parking lot is balanced. Specifically, the following steps are included: The path load index, path conflict risk coefficient and corresponding load conflict coefficient of several departure paths generated in the past period of time are collected and calibrated as and x represents the path load index, path conflict risk coefficient and the number of the corresponding load conflict coefficient of each departure path generated in the past period of time, x=3, 4, 5, ..., d, d is a positive integer, and the data collected in the past period of time is formed into a historical data set; Select the multivariate regression model as the load conflict model, and train it through the historical data set to determine the value of the regression coefficient according to the formula: Where β0, β1 and β2 are regression coefficients; By minimizing the error between the predicted value and the actual value, the regression coefficient is optimized, and finally the values ​​of the regression coefficients β0, β1 and β2 are determined; Using the final regression coefficient, the load conflict model is constructed and the path load index PLI of each departure path generated in real time is input. i and path conflict risk coefficient PCRC i , generate the load conflict coefficient LCC of each departure path in real time i ; Calculate the standard deviation of the load conflict coefficient of each departure path. The specific calculation formula is as follows: In the formula, σ LCC is the standard deviation of the load conflict coefficient of each departure path; Determine the standard deviation threshold of the load conflict coefficient for each pre-set departure path After determining, the standard deviation σ of the load conflict coefficient of each departure path LCC Compare and judge whether the load distribution of the parking lot's departure path is balanced according to the comparison results. The specific comparison and analysis is as follows: like The parking lot’s departure paths have a balanced load distribution; like The load distribution on the departure paths of the parking lot is uneven.

6. The parking order analysis method based on big data according to claim 5 is characterized in that: When it is determined that the load distribution of the departure path of the parking lot is unbalanced, a preset load conflict coefficient threshold interval [LCC min , LCC max ] and compare the pre-set load conflict coefficient threshold interval with the load conflict coefficient LCC of each departure path i Compare and divide each departure path into high-conflict path, normal path and low-load path according to the comparison results. The specific comparison analysis and division are as follows: If LCC i <LCC min , dividing the departure path into a low-load path; If LCC min ≤LCC i ≤LCC max , divide the departure path into a normal path; If LCC i >LCC max , classifying the departure path as a high-conflict path.

7. The parking order analysis method based on big data according to claim 6 is characterized in that: According to the classification results of each departure path, corresponding optimization scheduling measures are taken for different types of paths, specifically: For the paths that are divided into high-conflict paths, the specific measures taken are: implement diversion and control strategies, dynamically adjust the guidance information of the path guidance screen, guide some vehicles to choose low-load paths, and optimize the timing of signal lights to increase the traffic efficiency of the path; For the paths that are classified as normal paths, the specific measures taken are: maintain the existing scheduling strategy and guidance measures, regularly monitor the changes in its load conflict coefficient, ensure that the path load is in a stable state, and no additional intervention measures are required to ensure the smooth operation of the overall parking lot resources; For paths that are divided into low-load paths, the specific measures taken are: give priority to guiding more vehicles into this path, achieve reasonable vehicle diversion through path guidance screens and dynamic traffic distribution, and at the same time reduce its idle resources and optimize the overall path utilization of the parking lot.

8. A parking order analysis system based on big data, used to implement the parking order analysis method based on big data as described in any one of claims 1 to 7, characterized in that: It includes a departure path network construction module, a path dynamic information analysis module, a load conflict model generation module, a path load classification module, a path optimization scheduling module, and a dynamic update scheduling module; The departure path network construction module determines the departure path of the parking lot based on the parking order information in the parking lot, including the order generation time, end time and parking space location, combined with the overall layout of the parking lot, the historical vehicle driving trajectory and the exit distribution information, and assigns a unique path number and corresponding label to each departure path to build the departure path network model of the parking lot; The path dynamic information analysis module obtains the path dynamic information of each departure path in the parking lot in real time, analyzes it after acquisition, and generates the path load index and path conflict risk coefficient of each departure path respectively; The load conflict model generation module constructs a load conflict model based on the path load index and path conflict risk coefficient of each generated departure path, generates the load conflict coefficient of each departure path, and analyzes it after generation to determine whether the load distribution of the departure paths of the parking lot is balanced; The path load classification module determines a preset load conflict coefficient threshold interval when it is determined that the load distribution of the departure paths of the parking lot is uneven, and compares the preset load conflict coefficient threshold interval with the load conflict coefficient of each departure path, and divides each departure path into a high-conflict path, a normal path, and a low-load path according to the comparison result; The path optimization and scheduling module takes corresponding optimization and scheduling measures for different categories of paths according to the classification results of each departure path; Dynamically update the scheduling module, continuously update the path dynamic information and load conflict model, dynamically adjust the scheduling strategy according to real-time data, and ensure the load distribution balance and resource utilization efficiency of the departure path.

Images