Valet parking order prediction method and system based on fusion of multi-source data

Abstract

The present invention discloses a valet parking order prediction method and system based on the fusion of multi-source data, which relates to the field of urban traffic management technology. Firstly, a target area is evenly divided into a number of square areas, service resource information and parking space resource information data of parking service personnel are collected, and a parking resource shortage index Zkf is obtained by fitting after analysis, and a corresponding parking resource shortage signal is generated and sent to an order prediction module. At the same time, the driving roads in the target area are divided into independent sections, and features are extracted to obtain traffic flow information data. After the congestion degree is analyzed, the regional congestion index Zjt is obtained. Subsequently, based on relevant parking resource shortage signals and relevant traffic congestion degree signals, a comprehensive prediction of order demand is performed, and a corresponding comprehensive prediction order demand signal is generated and sent to an early warning feedback module. Finally, an order demand early warning signal of the corresponding level is generated and displayed.

Description

Technical Field

[0001] The present invention relates to the technical field of urban traffic management, and in particular to a method and system for predicting valet parking orders based on fusion of multi-source data. Background Art

[0002] With the acceleration of urbanization, traffic congestion and parking difficulties are becoming increasingly serious. This phenomenon has attracted widespread attention from all walks of life. As an important part of urban construction, urban traffic management has gradually shifted from traditional traffic flow management to the application of intelligent traffic management systems to improve traffic efficiency and user experience. In the field of urban traffic management, valet parking services, as an emerging parking solution, are gradually gaining favor among consumers. The valet parking order prediction system integrates and analyzes multi-source data to predict the demand for valet parking orders in order to improve service efficiency and better meet the growing parking needs of urban residents.

[0003] In the current prediction of the demand for valet parking orders, the traditional model often finds it difficult to respond to users' demand for valet parking in a timely manner, resulting in insufficient and wasted parking resources and a decline in the overall service quality. For example, it is difficult to predict the demand for valet parking orders in shopping malls, business districts and stadiums in many cities in real time. Due to the failure to formulate corresponding measures in a timely manner according to the demand for predicted orders, there is a shortage of parking space resources and service resources during peak hours, while there is a large amount of resource waste during non-peak hours. This situation not only affects the user's parking experience, but also leads to a vicious cycle of traffic congestion. In addition, there is a lack of effective multi-source data analysis and prediction methods, which makes it difficult to grasp the real-time traffic flow and parking resources, resulting in low efficiency of valet parking services. In response to this situation, a valet parking order prediction system based on the fusion of multi-source data can accurately predict the order demand through real-time data collection and analysis, and formulate corresponding measures to alleviate the problem of insufficient and wasted parking resources, and improve the overall valet parking service quality. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a valet parking order prediction method and system based on the fusion of multi-source data, which solves the problems in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a valet parking order prediction system based on the fusion of multi-source data, including a data acquisition module, a resource analysis module, a traffic monitoring module, a congestion analysis module, an order prediction module and an early warning feedback module;

[0006] The data collection module is used to evenly divide the target area into a number of square areas, and connect to the data source interface of the valet parking software through the application programming interface API, collect the service resource information of the parking service personnel in the target area, obtain the service resource information data, and collect the parking space resource information in the target area, and obtain the parking space resource information data;

[0007] The resource analysis module is used to analyze the service resource information data and the parking space resource information data, obtain the parking resource shortage index Zkf by fitting, generate relevant parking resource shortage signals after comparative analysis, and send them to the order prediction module;

[0008] The traffic monitoring module is used to divide the driving roads in the target area into several independent sections, monitor the traffic flow conditions of the several independent sections through multiple groups of traffic cameras, and obtain the traffic flow information data of each section after feature extraction;

[0009] The congestion analysis module is used to receive the traffic flow information data of each road section in the target area, analyze the congestion degree in the target area, obtain the regional congestion index Zjt, and generate the relevant traffic congestion degree signal after feature extraction and comparative analysis, and send it to the order prediction module;

[0010] The order prediction module is used to perform order demand comprehensive prediction and analysis processing on the received relevant parking resource shortage signal and relevant traffic congestion degree signal, generate relevant comprehensive prediction order demand signal, and send it to the early warning feedback module;

[0011] The early warning feedback module is used to perform early warning control analysis and processing on the received relevant comprehensive forecast order demand signals, generate order demand early warning signals of corresponding levels, and display and explain them.

[0012] Preferably, the data collection module includes an area division unit, a service resource collection unit and a parking space resource collection unit;

[0013] The area division unit is used to select an area with parking demand, and use geographic information system technology to define the element characteristics of the selected area, the element characteristic definition includes the length, width, geographical location and functional category of the area, and after the element characteristic definition, it is used as the target area, and the target area is evenly divided into a number of square areas;

[0014] The service resource collection unit is used to connect to the data source interface of the valet parking software through the application programming interface API, collect service resource information of parking service personnel in a number of square areas, and obtain service resource information data, wherein the service resource information data includes the number of idle service personnel Nkx, the service personnel density Mfw and the average order acceptance time Tjd of the service personnel in each square area;

[0015] The parking space resource collection unit is used to utilize multiple groups of deployed traffic cameras and combine video surveillance technology to collect parking space resource information on both sides of the road and in parking garages in several square areas to obtain parking space resource information data, wherein the parking space resource information data includes the number of vacant parking spaces Ncw and the average parking time Ttf of vehicles in each square area.

[0016] Preferably, the resource analysis module includes a service resource analysis unit, a parking space resource analysis unit and a resource scarcity assessment unit;

[0017] The service resource analysis unit is used to analyze the parking space resource information data in each square area, and obtain the service resource shortage index Zfw after dimensionless processing. The service resource shortage index Zfw is calculated and obtained by the following formula:

[0018] ;

[0019] In the formula, It is represented as the average order-taking time of service personnel in the i-th square area, It is represented by the number of idle service personnel in the i-th square area, It is represented as the service personnel density in the i-th square area, i = 1, 2, 3, ..., n, and n represents the number of square areas;

[0020] The parking space resource analysis unit is used to analyze the parking space resource information data in each square area, and obtain the parking space resource scarcity index Zcw after dimensionless processing. The parking space resource scarcity index Zcw is calculated and obtained by the following formula:

[0021] ;

[0022] In the formula, It is represented by the number of vacant parking spaces in the i-th square area, It is represented as the average parking time of vehicles in the i-th square area, i=1, 2, 3, ..., n, and n represents the number of square areas.

[0023] Preferably, the resource shortage evaluation unit is used to associate the service resource shortage index Zfw with the parking space resource shortage index Zcw, and obtain the parking resource shortage index Zkf after dimensionless processing. The parking resource shortage index Zkf is calculated and obtained by the following formula:

[0024] ;

[0025] In the formula, and They are respectively represented as the weight values ​​of the service resource scarcity index Zfw and the parking space resource scarcity index Zcw, and A is represented as the correction constant;

[0026] By presetting a resource shortage reference threshold T and comparing and analyzing the resource shortage reference threshold T with the parking resource shortage index Zkf in the target area, a relevant parking resource shortage signal is generated. The specific process is as follows;

[0027] When the parking resource shortage index Zkf is greater than the preset resource shortage reference threshold T, a "parking resource severe shortage signal" is generated; when the parking resource shortage index Zkf is less than or equal to the preset resource shortage reference threshold T, a "parking resource mild shortage signal" is generated, and the generated "parking resource severe shortage signal" and "parking resource mild shortage signal" are sent to the order prediction module;

[0028] The relevant parking resource shortage signal includes a “parking resource severe shortage signal” and a “parking resource mild shortage signal”.

[0029] Preferably, the traffic monitoring module includes a road segment division unit and a monitoring unit;

[0030] The road segment division unit is based on a road network model and is used to perform feature recognition on roads in the target area. Feature recognition includes spatial location, road type and road length of the road. After feature recognition, the roads in the target area are divided into a number of independent target road segments.

[0031] The monitoring unit is used to monitor the traffic flow conditions of several independent target sections through multiple groups of traffic cameras, and obtain traffic flow information data after feature extraction. The traffic flow information data includes the vehicle distance and vehicle speed of each target section, and after preprocessing with a statistical averaging algorithm, the average vehicle distance Dcs and the average vehicle speed Vcs of each target section are obtained respectively.

[0032] Preferably, the congestion analysis module includes a congestion pre-analysis unit and a peak congestion determination unit;

[0033] The congestion pre-analysis unit is used to analyze the pre-processed traffic flow information data, and after dimensionless processing, fit to obtain m groups of road section congestion index Zdd. The calculation method of each group of the m groups of road section congestion index Zdd is the same. Take one group of road section congestion index Zdd as an example: the road section congestion index of the jth target road section is , which is obtained by the following formula:

[0034] ;

[0035] In the formula, It is expressed as the average distance between vehicles on the jth target section. It is represented as the average speed of vehicles traveling on the jth target road section, where and They are all expressed as weight values, j belongs to 1, 2, 3, ..., m, and B is expressed as the second correction constant.

[0036] Preferably, the peak congestion determination unit is used to extract features from the m groups of acquired road congestion indexes Zdd to obtain the maximum road congestion index Zdd of each target road segment. max and the minimum road congestion index Zdd min The obtained m groups of road section congestion indexes Zdd are processed by a statistical averaging algorithm to obtain the regional congestion index Zjt;

[0037] According to the maximum road congestion index Zdd of each target road section max and the minimum road congestion index Zdd min , fitting to obtain reasonable traffic flow range ;

[0038] ;

[0039] Among them, k is a constant, and the specific value is set by the user;

[0040] By combining the regional congestion index Zjt with the reasonable traffic flow range Comparative analysis is performed to determine whether the traffic flow in the current target area is in a peak congestion state, and to generate relevant traffic congestion degree signals. The specific process is as follows:

[0041] If the regional congestion index Zjt exceeds the reasonable traffic flow range , it indicates that the traffic flow in the current target area is in a peak congestion state, and a "severe traffic congestion signal" is generated; if the congestion index Zjt of the area does not exceed the reasonable traffic flow range When , it indicates that the traffic flow in the current target area is not in a peak congestion state, a "traffic light congestion signal" is generated, and both the "traffic heavy congestion signal" and the "traffic light congestion signal" are sent to the order prediction module;

[0042] The relevant traffic congestion degree signals include "severe traffic congestion signal" and "mild traffic congestion signal".

[0043] Preferably, the order prediction module is used to perform order demand comprehensive prediction analysis processing on the received relevant parking resource shortage signal and relevant traffic congestion degree signal to generate relevant comprehensive prediction order demand signal. The specific process is as follows:

[0044] A set X is established according to the relevant parking resource shortage signals, a "parking resource mild shortage signal" is marked as element a1, a "parking resource severe shortage signal" is marked as element a2, and element a1∈set X, element a2∈set X;

[0045] According to the relevant traffic congestion degree signals, a set Y is established, "traffic light congestion signal" is marked as element b1, "traffic heavy congestion signal" is marked as element b2, and element b1∈set Y, element b2∈set Y;

[0046] Performing a union process on the set X and the set Y, if X∪Y={a2, b2}, then generating a “comprehensively predicted heavy order demand signal”, if X∪Y={a1, b2} or {a2, b1}, then generating a “comprehensively predicted moderate order demand signal”, if X∪Y={a1, b1}, then generating a “comprehensively predicted light order demand signal”, and sending the “comprehensively predicted heavy order demand signal”, “comprehensively predicted moderate order demand signal” and “comprehensively predicted light order demand signal” to the early warning feedback module;

[0047] The relevant comprehensive forecast order demand signal includes "comprehensive forecast heavy order demand signal", "comprehensive forecast medium order demand signal" and "comprehensive forecast light order demand signal".

[0048] Preferably, the early warning feedback module is used to perform early warning control analysis and processing on the received relevant comprehensive order demand signals, generate order demand early warning signals of corresponding levels, and display and explain them. The specific process is as follows;

[0049] When a "comprehensive prediction heavy order demand signal" is received, a first-level order demand warning signal is generated and sent to the display terminal in the form of a text message "It is predicted that the valet parking order in the current target area is in a heavy demand state", and a red flashing indicator light is lit;

[0050] When the "comprehensive forecast medium order demand signal" is received, a secondary order demand warning signal is generated and sent to the display terminal in the form of a text message "It is predicted that the valet parking order in the current target area is in a medium demand state", and a yellow flashing indicator light is lit;

[0051] When a "comprehensively predicted light order demand signal" is received, a third-level order demand warning signal is generated and sent to the display terminal for display in the form of text "It is predicted that the valet parking orders in the current target area are in a light demand state", and a blue flashing indicator light is turned on.

[0052] The valet parking order prediction method based on the fusion of multi-source data includes the following steps:

[0053] Step 1: First, the target area is evenly divided into several square areas, and the data source interface of the valet parking software is connected through the application programming interface API to collect the service resource information of the parking service personnel in the target area to obtain the service resource information data, and the parking space resource information in the target area is collected to obtain the parking space resource information data; based on the analysis of the service resource information data and the parking space resource information data, the parking resource scarcity index Zkf is obtained by fitting, and after comparative analysis, the relevant parking resource scarcity signal is generated and sent to the order prediction module;

[0054] Step 2: Secondly, divide the driving roads in the target area into several independent sections, monitor the traffic flow of several independent sections through multiple groups of traffic cameras, and obtain the traffic flow information data of each section after feature extraction; and analyze the congestion level in the target area to obtain the regional congestion index Zjt. After feature extraction and comparative analysis, generate relevant traffic congestion level signals and send them to the order prediction module;

[0055] Step 3: Then, the received parking resource shortage signal and traffic congestion signal are processed for comprehensive order demand forecasting and analysis, and a comprehensive forecast order demand signal is generated, and sent to the early warning feedback module;

[0056] Step 4: Finally, the received relevant comprehensive forecast order demand signals are analyzed and processed for early warning control, order demand early warning signals of corresponding levels are generated, and displayed and explained.

[0057] The present invention provides a valet parking order prediction method and system based on the fusion of multi-source data, which has the following beneficial effects:

[0058] (1) Through real-time data analysis and comprehensive prediction, accurate prediction of valet parking order demand is achieved. This process not only reduces the waste of resources caused by the failure to respond to valet parking order demand in a timely manner, but also can dynamically adjust the service strategy according to real-time demand. At the same time, the effective early warning function enables service personnel to prepare in advance, further improving service quality and user satisfaction. Through scientific data analysis and real-time order demand prediction, users will enjoy better quality and more efficient valet parking services, significantly improving the overall valet parking service quality.

[0059] (2) Through the traffic monitoring module and the congestion analysis module, the road traffic volume is monitored in real time to obtain the regional congestion index Zjt, which can effectively judge the congestion situation. By integrating the parking resource shortage signal and the traffic congestion degree signal, the order prediction module can comprehensively analyze the order demand, predict in advance and formulate corresponding measures to avoid the aggravation of traffic congestion during peak hours. This intelligent early warning mechanism not only improves the efficiency of valet parking services, but also provides important data support for urban traffic management, which helps to alleviate urban traffic congestion problems.

[0060] (3) By evenly dividing the target area and collecting data, it is possible to obtain information about parking service personnel and parking space resources in real time. This systematic data analysis not only effectively solves the problem of insufficient resource supply and waste in the traditional model, but also can respond to user needs in real time and improve the management efficiency of parking resources. By generating the parking resource scarcity index Zkf, the system can promptly identify resource shortages and thus optimize parking resource scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a block diagram of a valet parking order prediction system based on fusion of multi-source data of the present invention;

[0062] Figure 2 It is a schematic flow chart of the method for predicting valet parking orders based on the fusion of multi-source data of the present invention. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0064] Example 1

[0065] See also Figure 1, the present invention provides a valet parking order prediction system based on the fusion of multi-source data, including a data acquisition module, a resource analysis module, a traffic monitoring module, a congestion analysis module, an order prediction module and an early warning feedback module;

[0066] The data collection module is used to evenly divide the target area into a number of square areas, and connect to the data source interface of the valet parking software through the application programming interface API, collect the service resource information of the parking service personnel in the target area, obtain the service resource information data, and collect the parking space resource information in the target area, and obtain the parking space resource information data;

[0067] The resource analysis module is used to analyze the service resource information data and the parking space resource information data, obtain the parking resource shortage index Zkf by fitting, generate relevant parking resource shortage signals after comparative analysis, and send them to the order prediction module;

[0068] The traffic monitoring module is used to divide the driving roads in the target area into several independent sections, monitor the traffic flow conditions of the several independent sections through multiple groups of traffic cameras, and obtain the traffic flow information data of each section after feature extraction;

[0069] The congestion analysis module is used to receive the traffic flow information data of each road section in the target area, analyze the congestion degree in the target area, obtain the regional congestion index Zjt, and generate the relevant traffic congestion degree signal after feature extraction and comparative analysis, and send it to the order prediction module;

[0070] The order prediction module is used to perform order demand comprehensive prediction and analysis processing on the received relevant parking resource shortage signal and relevant traffic congestion degree signal, generate relevant comprehensive prediction order demand signal, and send it to the early warning feedback module;

[0071] The early warning feedback module is used to perform early warning control analysis and processing on the received relevant comprehensive forecast order demand signals, generate order demand early warning signals of corresponding levels, and display and explain them.

[0072] In this embodiment, by integrating real-time data collection and intelligent analysis, the efficiency and service quality of urban parking management are improved, and the problems of insufficient parking resources and waste of resources existing in the traditional mode are effectively dealt with. First, the target area is divided into several square areas, and the application programming interface API is used to collect parking service personnel and parking space resource information in real time to generate a parking resource shortage index Zkf, so as to achieve accurate prediction of parking demand, thereby avoiding the embarrassing situation of insufficient parking resource supply during peak hours. At the same time, the effective combination of flow monitoring and congestion analysis modules enables the system to monitor the traffic flow changes in the target area in real time and identify traffic congestion in time. The order prediction module achieves accurate prediction of order demand by comprehensively analyzing parking resource shortage signals and traffic congestion degree signals, and improves the response speed and overall user experience of valet parking services. This series of innovative measures not only achieves accurate prediction of valet parking order demand, but also effectively alleviates the prominent problems of difficulty in parking, waste of resources and traffic congestion, but also provides a practical solution for the intelligent and refined management of valet parking services, improves the overall valet parking service quality, and promotes the modernization process of urban traffic management.

[0073] Example 2

[0074] Please refer to Figure 1 ,Specifically: the data acquisition module includes an area division unit, a service resource acquisition unit and a parking space resource acquisition unit;

[0075] The area division unit is used to select an area with parking demand, and to define the element characteristics of the selected area through geographic information system technology, wherein the element characteristic definition includes the length, width, geographical location and functional category of the area, and after the element characteristic definition, the target area is used as the target area, and the target area is evenly divided into a plurality of square areas, wherein dividing the target area into a plurality of square areas can realize refined data collection and resource management;

[0076] The service resource collection unit is used to connect to the data source interface of the valet parking software through the application programming interface API, collect service resource information of parking service personnel in a number of square areas, and obtain service resource information data, wherein the service resource information data includes the number of idle service personnel Nkx, the service personnel density Mfw and the average order acceptance time Tjd of the service personnel in each square area;

[0077] The parking space resource collection unit is used to utilize multiple groups of deployed traffic cameras and combine video surveillance technology to collect parking space resource information on both sides of the road and in parking garages in several square areas to obtain parking space resource information data, wherein the parking space resource information data includes the number of vacant parking spaces Ncw and the average parking time Ttf of vehicles in each square area.

[0078] In this embodiment, the response capability and service efficiency of the valet parking order prediction system are improved through the close cooperation of the area division unit, the service resource collection unit and the parking space resource collection unit. The area division unit uses geographic information system technology to accurately divide the area with parking demand, so that the characteristics of the target area are clearer, providing a solid foundation for subsequent data collection. The service resource collection unit obtains the number of idle service personnel Nkx, service personnel density Mfw and average order acceptance time Tjd of service personnel in each square area in real time through the API, ensuring dynamic monitoring and rapid deployment of service resources. At the same time, the parking space resource collection unit combines multiple groups of traffic cameras to monitor and collect information on parking spaces in real time, so that the system can timely identify the number of vacant parking spaces Ncw and the average parking time Ttf of vehicles in each square area, solving the common problems of waste and insufficient utilization of parking space resources in the traditional mode. This comprehensive data collection capability enables the system to provide more accurate data support for urban parking problems.

[0079] Example 3

[0080] Please refer to Figure 1 ,Specifically: the resource analysis module includes a service resource analysis unit, a parking space resource analysis unit and a resource scarcity assessment unit;

[0081] The service resource analysis unit is used to analyze the parking space resource information data in each square area, and obtain the service resource shortage index Zfw after dimensionless processing. The service resource shortage index Zfw is calculated and obtained by the following formula:

[0082] ;

[0083] In the formula, It is represented as the average order-taking time of service personnel in the i-th square area, It is represented by the number of idle service personnel in the i-th square area, It is represented as the service personnel density in the i-th square area, i = 1, 2, 3, ..., n, and n represents the number of square areas;

[0084] The parking space resource analysis unit is used to analyze the parking space resource information data in each square area, and obtain the parking space resource scarcity index Zcw after dimensionless processing. The parking space resource scarcity index Zcw is calculated and obtained by the following formula:

[0085] ;

[0086] In the formula, It is represented by the number of vacant parking spaces in the i-th square area, It is represented as the average parking time of vehicles in the i-th square area, i=1, 2, 3, ..., n, and n represents the number of square areas.

[0087] Specifically, the resource shortage evaluation unit is used to associate the service resource shortage index Zfw with the parking space resource shortage index Zcw, and obtain the parking resource shortage index Zkf after dimensionless processing. The parking resource shortage index Zkf is calculated and obtained by the following formula:

[0088] ;

[0089] In the formula, and They are respectively represented as the weight values ​​of the service resource scarcity index Zfw and the parking space resource scarcity index Zcw, and A is represented as the correction constant;

[0090] By presetting a resource shortage reference threshold T and comparing and analyzing the resource shortage reference threshold T with the parking resource shortage index Zkf in the target area, a relevant parking resource shortage signal is generated. The specific process is as follows;

[0091] When the parking resource shortage index Zkf is greater than the preset resource shortage reference threshold T, a "parking resource severe shortage signal" is generated; when the parking resource shortage index Zkf is less than or equal to the preset resource shortage reference threshold T, a "parking resource mild shortage signal" is generated, and the generated "parking resource severe shortage signal" and "parking resource mild shortage signal" are sent to the order prediction module;

[0092] The relevant parking resource shortage signal includes a “parking resource severe shortage signal” and a “parking resource mild shortage signal”.

[0093] In this embodiment, the management and allocation efficiency of parking resources is improved through the collaborative work of the service resource analysis unit, the parking space resource analysis unit and the resource shortage assessment unit. By performing a detailed analysis of the service resource and parking space resource information in each square area, the service resource shortage index Zfw and the parking space resource shortage index Zcw can be accurately calculated, and finally the parking resource shortage index Zkf is obtained. This process ensures real-time monitoring of the parking resource status and can timely identify the shortage or surplus of resources. Compared with the traditional prediction model, the severe and mild shortage states of parking resources can be accurately distinguished through the set resource shortage reference threshold T. This forward-looking resource assessment process not only optimizes the efficiency of valet parking services and improves user experience, but also provides a more intelligent solution for urban traffic management, which fully meets the urgent needs of modern cities in dealing with parking problems and traffic congestion challenges.

[0094] Example 4

[0095] Please refer to Figure 1 ,Specifically: the traffic monitoring module includes a road section division unit and a monitoring unit;

[0096] The road segment division unit is based on a road network model and is used to perform feature recognition on roads in the target area, wherein the feature recognition includes the spatial position, road type and road length of the road. After feature recognition, the roads in the target area are divided into a plurality of independent target road segments, wherein dividing the roads in the target area into a plurality of independent target road segments helps to achieve more accurate traffic flow monitoring and analysis;

[0097] It should be noted that the road network model is a digital representation for simulating and analyzing road layout and its functions, established by integrating geographic information system data and traffic engineering principles. It is based on the spatial location, type and length characteristics of actual roads, and uses geographic coordinates and traffic flow data to construct a dynamic network structure to facilitate flow monitoring, congestion analysis and traffic management decisions.

[0098] The monitoring unit is used to monitor the traffic flow conditions of several independent target sections through multiple groups of traffic cameras, and obtain traffic flow information data after feature extraction. The traffic flow information data includes the vehicle distance and vehicle speed of each target section, and after preprocessing with a statistical averaging algorithm, the average vehicle distance Dcs and the average vehicle speed Vcs of each target section are obtained respectively.

[0099] In this embodiment, the real-time monitoring and analysis capabilities of traffic flow in the target area are improved through the efficient integration of the road segmentation unit and the monitoring unit. The road segmentation unit performs feature recognition based on the road network model, accurately identifies the spatial position, type and length of each road, and effectively divides the roads in the target area into multiple independent target sections. This precise division provides a solid foundation for subsequent traffic monitoring. The monitoring unit uses multiple groups of traffic cameras to monitor the traffic flow of each independent target section in real time, extracts the vehicle distance and driving speed data, and uses statistical algorithms to perform mean preprocessing to obtain the average vehicle distance Dcs and the average speed Vcs of each target section. This comprehensive monitoring capability not only realizes the real-time grasp of traffic flow, helps to identify and respond to congestion in a timely manner, but also provides accurate data support for the order prediction module, optimizes the deployment strategy of valet parking services, effectively alleviates the problem of lack of real-time feedback on traffic flow and parking demand in the traditional mode, and further improves the intelligence and efficiency of urban traffic management.

[0100] Example 5

[0101] Please refer to Figure 1,Specifically: the congestion analysis module includes a congestion pre-analysis unit and a peak congestion determination unit;

[0102] The congestion pre-analysis unit is used to analyze the pre-processed traffic flow information data, and after dimensionless processing, fit to obtain m groups of road section congestion index Zdd. The calculation method of each group of the m groups of road section congestion index Zdd is the same. Take one group of road section congestion index Zdd as an example: the road section congestion index of the jth target road section is , which is obtained by the following formula:

[0103] ;

[0104] In the formula, It is expressed as the average distance between vehicles on the jth target section. It is represented as the average speed of vehicles traveling on the jth target road section, where and They are all expressed as weight values, j belongs to 1, 2, 3, ..., m, and B is expressed as the second correction constant.

[0105] Specifically, the peak congestion determination unit is used to extract features from the m groups of road congestion indexes Zdd obtained to obtain the maximum road congestion index Zdd of each target road segment. max and the minimum road congestion index Zdd min The obtained m groups of road section congestion indexes Zdd are processed by a statistical averaging algorithm to obtain the regional congestion index Zjt;

[0106] According to the maximum road congestion index Zdd of each target road section max and the minimum road congestion index Zdd min , fitting to obtain reasonable traffic flow range ;

[0107] ;

[0108] Among them, k is a constant, and the specific value is set by the user;

[0109] By combining the regional congestion index Zjt with the reasonable traffic flow range Comparative analysis is performed to determine whether the traffic flow in the current target area is in a peak congestion state, and to generate relevant traffic congestion degree signals. The specific process is as follows:

[0110] If the regional congestion index Zjt exceeds the reasonable traffic flow range , it indicates that the traffic flow in the current target area is in a peak congestion state, and a "severe traffic congestion signal" is generated; if the congestion index Zjt of the area does not exceed the reasonable traffic flow range When , it indicates that the traffic flow in the current target area is not in a peak congestion state, a "traffic light congestion signal" is generated, and both the "traffic heavy congestion signal" and the "traffic light congestion signal" are sent to the order prediction module;

[0111] The relevant traffic congestion degree signals include "severe traffic congestion signal" and "mild traffic congestion signal".

[0112] In this embodiment, the congestion pre-analysis unit and the peak congestion determination unit are closely combined to achieve accurate identification and real-time monitoring of the traffic congestion status in the target area. The congestion pre-analysis unit performs dimensionless processing on the pre-processed vehicle flow information data to fit the section congestion index Zdd of each target section, making the analysis process more scientific and accurate. The peak congestion determination unit extracts the maximum section congestion index Zdd in the target section. max and the minimum road congestion index Zdd min , and use statistical methods to calculate the regional congestion index Zjt, which provides a reliable basis for judging whether the traffic flow is in a peak congestion state. When the regional congestion index exceeds the reasonable traffic flow range When a car arrives at a certain place, it can quickly generate a "signal of severe traffic congestion", otherwise it will generate a "signal of light traffic congestion", and feed these signals back to the order prediction module. This efficient congestion analysis capability not only helps to timely identify and respond to traffic congestion problems, improve the response speed and allocation efficiency of valet parking services, but also solves the problem of slow response to traffic flow changes in the traditional model, thereby effectively reducing the waste of parking resources and the vicious cycle of traffic congestion, reflecting the urgent need for real-time data analysis and response capabilities.

[0113] Example 6

[0114] Please refer to Figure 1 Specifically: the order prediction module is used to perform order demand comprehensive prediction analysis on the received relevant parking resource shortage signal and relevant traffic congestion degree signal to generate relevant comprehensive prediction order demand signal. The specific process is as follows:

[0115] A set X is established according to the relevant parking resource shortage signals, a "parking resource mild shortage signal" is marked as element a1, a "parking resource severe shortage signal" is marked as element a2, and element a1∈set X, element a2∈set X;

[0116] According to the relevant traffic congestion degree signals, a set Y is established, "traffic light congestion signal" is marked as element b1, "traffic heavy congestion signal" is marked as element b2, and element b1∈set Y, element b2∈set Y;

[0117] Performing a union process on the set X and the set Y, if X∪Y={a2, b2}, then generating a “comprehensively predicted heavy order demand signal”, if X∪Y={a1, b2} or {a2, b1}, then generating a “comprehensively predicted moderate order demand signal”, if X∪Y={a1, b1}, then generating a “comprehensively predicted light order demand signal”, and sending the “comprehensively predicted heavy order demand signal”, “comprehensively predicted moderate order demand signal” and “comprehensively predicted light order demand signal” to the early warning feedback module;

[0118] The relevant comprehensive forecast order demand signal includes "comprehensive forecast heavy order demand signal", "comprehensive forecast medium order demand signal" and "comprehensive forecast light order demand signal".

[0119] Specifically: the early warning feedback module is used to perform early warning control analysis and processing on the received relevant comprehensive order demand signals, generate order demand early warning signals of corresponding levels, and display and explain them. The specific process is as follows;

[0120] When a "comprehensive prediction heavy order demand signal" is received, a first-level order demand warning signal is generated and sent to the display terminal in the form of a text message "It is predicted that the valet parking order in the current target area is in a heavy demand state", and a red flashing indicator light is lit;

[0121] When the "comprehensive forecast medium order demand signal" is received, a secondary order demand warning signal is generated and sent to the display terminal in the form of a text message "It is predicted that the valet parking order in the current target area is in a medium demand state", and a yellow flashing indicator light is lit;

[0122] When a "comprehensively predicted light order demand signal" is received, a third-level order demand warning signal is generated and sent to the display terminal for display in the form of text "It is predicted that the valet parking orders in the current target area are in a light demand state", and a blue flashing indicator light is turned on.

[0123] In this embodiment, by comprehensively analyzing the relevant parking resource shortage signals and the relevant traffic congestion degree signals, the accurate prediction of the valet parking order demand is achieved, and the system's responsiveness and service efficiency are improved. By constructing set X and set Y and performing union processing, the heavy, medium and light states of the current parking demand can be clearly identified, thereby generating a corresponding comprehensive predicted order demand signal. This mechanism effectively solves the pain point of the traditional valet parking mode that it is difficult to respond to user needs in real time, so that service personnel can prepare in advance according to the predicted information, optimize resource allocation, avoid service shortages during peak periods and waste of resources during low peak periods. At the same time, the early warning feedback module generates different levels of order demand early warning signals in real time and provides clear information prompts on the display end to ensure that relevant personnel can quickly understand the current demand status and take corresponding measures. This real-time prediction and feedback mechanism not only improves the user's parking experience, but also fundamentally alleviates the problem of urban traffic congestion, fully reflecting the importance and necessity of modern intelligent traffic management systems in dealing with complex urban traffic environments.

[0124] Example 7

[0125] Please refer to Figure 1 and Figure 2 ,Specifically: A valet parking order prediction method based on fusion of multi-source data, comprises the following steps;

[0126] Step 1: First, the target area is evenly divided into several square areas, and the data source interface of the valet parking software is connected through the application programming interface API to collect the service resource information of the parking service personnel in the target area to obtain the service resource information data, and the parking space resource information in the target area is collected to obtain the parking space resource information data; based on the analysis of the service resource information data and the parking space resource information data, the parking resource scarcity index Zkf is obtained by fitting, and after comparative analysis, the relevant parking resource scarcity signal is generated and sent to the order prediction module;

[0127] Step 2: Secondly, divide the driving roads in the target area into several independent sections, monitor the traffic flow of several independent sections through multiple groups of traffic cameras, and obtain the traffic flow information data of each section after feature extraction; and analyze the congestion level in the target area to obtain the regional congestion index Zjt. After feature extraction and comparative analysis, generate relevant traffic congestion level signals and send them to the order prediction module;

[0128] Step 3: Then, the received parking resource shortage signal and traffic congestion signal are processed for comprehensive order demand forecasting and analysis, and a comprehensive forecast order demand signal is generated, and sent to the early warning feedback module;

[0129] Step 4: Finally, the received relevant comprehensive forecast order demand signals are analyzed and processed for early warning control, order demand early warning signals of corresponding levels are generated, and displayed and explained.

[0130] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A valet parking order prediction system based on the fusion of multi-source data, characterized by: It includes data collection module, resource analysis module, traffic monitoring module, congestion analysis module, order prediction module and early warning feedback module; The data collection module is used to evenly divide the target area into a number of square areas, and connect to the data source interface of the valet parking software through the application programming interface API, collect the service resource information of the parking service personnel in the target area, obtain the service resource information data, and collect the parking space resource information in the target area, and obtain the parking space resource information data; The data collection module includes an area division unit, a service resource collection unit and a parking space resource collection unit; The area division unit is used to select an area with parking demand, and use geographic information system technology to define the element characteristics of the selected area, the element characteristic definition includes the length, width, geographical location and functional category of the area, and after the element characteristic definition, it is used as the target area, and the target area is evenly divided into a number of square areas; The service resource collection unit is used to connect to the data source interface of the valet parking software through the application programming interface API, collect service resource information of parking service personnel in a number of square areas, and obtain service resource information data, wherein the service resource information data includes the number of idle service personnel Nkx, the service personnel density Mfw and the average order acceptance time Tjd of the service personnel in each square area; The parking space resource collection unit is used to collect parking space resource information on both sides of the road and the parking garage in a number of square areas by using multiple groups of deployed traffic cameras and combining video surveillance technology to obtain parking space resource information data, wherein the parking space resource information data includes the number of vacant parking spaces Ncw and the average parking time Ttf of vehicles in each square area; The resource analysis module is used to analyze the service resource information data and the parking space resource information data, obtain the parking resource shortage index Zkf by fitting, generate relevant parking resource shortage signals after comparative analysis, and send them to the order prediction module; The resource analysis module includes a service resource analysis unit, a parking space resource analysis unit and a resource scarcity assessment unit; The service resource analysis unit is used to analyze the parking space resource information data in each square area, and obtain the service resource shortage index Zfw after dimensionless processing. The service resource shortage index Zfw is calculated and obtained by the following formula: ； In the formula, It is represented as the average order-taking time of service personnel in the i-th square area, It is represented by the number of idle service personnel in the i-th square area, It is represented as the service personnel density in the i-th square area, i = 1, 2, 3, ..., n, and n represents the number of square areas; The parking space resource analysis unit is used to analyze the parking space resource information data in each square area, and obtain the parking space resource scarcity index Zcw after dimensionless processing. The parking space resource scarcity index Zcw is calculated and obtained by the following formula: ； In the formula, It is represented by the number of vacant parking spaces in the i-th square area, It is represented as the average parking time of vehicles in the i-th square area, i = 1, 2, 3, ..., n, and n represents the number of square areas; The resource shortage evaluation unit is used to associate the service resource shortage index Zfw with the parking space resource shortage index Zcw, and obtain the parking resource shortage index Zkf after dimensionless processing. The parking resource shortage index Zkf is calculated and obtained by the following formula: ； In the formula, and They are respectively represented as the weight values ​​of the service resource scarcity index Zfw and the parking space resource scarcity index Zcw, and A is represented as the correction constant; By presetting a resource shortage reference threshold T and comparing and analyzing the resource shortage reference threshold T with the parking resource shortage index Zkf in the target area, a relevant parking resource shortage signal is generated. The specific process is as follows; When the parking resource shortage index Zkf is greater than the preset resource shortage reference threshold T, a "parking resource severe shortage signal" is generated; when the parking resource shortage index Zkf is less than or equal to the preset resource shortage reference threshold T, a "parking resource mild shortage signal" is generated, and the generated "parking resource severe shortage signal" and "parking resource mild shortage signal" are sent to the order prediction module; The relevant parking resource shortage signal includes a "parking resource severe shortage signal" and a "parking resource mild shortage signal"; The traffic monitoring module is used to divide the driving roads in the target area into several independent sections, monitor the traffic flow conditions of the several independent sections through multiple groups of traffic cameras, and obtain the traffic flow information data of each section after feature extraction; The congestion analysis module is used to receive the traffic flow information data of each road section in the target area, analyze the congestion degree in the target area, obtain the regional congestion index Zjt, and generate the relevant traffic congestion degree signal after feature extraction and comparative analysis, and send it to the order prediction module; The order prediction module is used to perform order demand comprehensive prediction and analysis processing on the received relevant parking resource shortage signal and relevant traffic congestion degree signal, generate relevant comprehensive prediction order demand signal, and send it to the early warning feedback module; The early warning feedback module is used to perform early warning control analysis and processing on the received relevant comprehensive forecast order demand signals, generate order demand early warning signals of corresponding levels, and display and explain them.

2. The valet parking order prediction system based on fusion of multi-source data according to claim 1 is characterized by: The traffic monitoring module includes a road segment division unit and a monitoring unit; The road segment division unit is based on a road network model and is used to perform feature recognition on roads in the target area. Feature recognition includes spatial location, road type and road length of the road. After feature recognition, the roads in the target area are divided into a number of independent target road segments. The monitoring unit is used to monitor the traffic flow conditions of several independent target sections through multiple groups of traffic cameras, and obtain traffic flow information data after feature extraction. The traffic flow information data includes the vehicle distance and vehicle speed of each target section, and after preprocessing with a statistical averaging algorithm, the average vehicle distance Dcs and the average vehicle speed Vcs of each target section are obtained respectively.

3. The valet parking order prediction system based on fusion of multi-source data according to claim 2 is characterized by: The congestion analysis module includes a congestion pre-analysis unit and a peak congestion determination unit; The congestion pre-analysis unit is used to analyze the pre-processed traffic flow information data, and after dimensionless processing, fit to obtain m groups of road section congestion index Zdd. The calculation method of each group of the m groups of road section congestion index Zdd is the same. Take one group of road section congestion index Zdd as an example: the road section congestion index of the jth target road section is , which is obtained by the following formula: ； In the formula, It is expressed as the average distance between vehicles on the jth target section. It is represented as the average speed of vehicles traveling on the jth target road section, where and They are all expressed as weight values, j belongs to 1, 2, 3, ..., m, and B is expressed as the second correction constant.

4. The valet parking order prediction system based on fusion of multi-source data according to claim 3 is characterized by: The peak congestion determination unit is used to extract features from the m groups of road congestion indexes Zdd to obtain the maximum road congestion index Zdd of each target road segment. max and the minimum road congestion index Zdd min The obtained m groups of road section congestion indexes Zdd are processed by a statistical averaging algorithm to obtain the regional congestion index Zjt; According to the maximum road congestion index Zdd of each target road section max and the minimum road congestion index Zdd min , fitting to obtain reasonable traffic flow range ; ； Among them, k is a constant, and the specific value is set by the user; By combining the regional congestion index Zjt with the reasonable traffic flow range Comparative analysis is performed to determine whether the traffic flow in the current target area is in a peak congestion state, and to generate relevant traffic congestion degree signals. The specific process is as follows: If the regional congestion index Zjt exceeds the reasonable traffic flow range , it indicates that the traffic flow in the current target area is in a peak congestion state, and a "severe traffic congestion signal" is generated; if the regional congestion index Zjt does not exceed the reasonable traffic flow range When , it indicates that the traffic flow in the current target area is not in a peak congestion state, a "traffic light congestion signal" is generated, and the "traffic heavy congestion signal" and the "traffic light congestion signal" are sent to the order prediction module; The relevant traffic congestion degree signals include "severe traffic congestion signal" and "mild traffic congestion signal".

5. The valet parking order prediction system based on fusion of multi-source data according to claim 1, characterized in that: The order prediction module is used to perform order demand comprehensive prediction analysis on the received relevant parking resource shortage signal and relevant traffic congestion degree signal to generate relevant comprehensive prediction order demand signal. The specific process is as follows: A set X is established according to the relevant parking resource shortage signals, "parking resource mild shortage signal" is marked as element a1, "parking resource severe shortage signal" is marked as element a2, and element a1∈set X, element a2∈set X; A set Y is established according to the relevant traffic congestion degree signals, "traffic light congestion signal" is marked as element b1, "traffic heavy congestion signal" is marked as element b2, and element b1∈set Y, element b2∈set Y; Performing a union process on the set X and the set Y, if X∪Y={a2, b2}, then generating a "comprehensively predicted heavy order demand signal", if X∪Y={a1, b2} or {a2, b1}, then generating a "comprehensively predicted moderate order demand signal", if X∪Y={a1, b1}, then generating a "comprehensively predicted light order demand signal", and sending the "comprehensively predicted heavy order demand signal", "comprehensively predicted moderate order demand signal" and "comprehensively predicted light order demand signal" to the early warning feedback module; The relevant comprehensive forecast order demand signal includes "comprehensive forecast heavy order demand signal", "comprehensive forecast medium order demand signal" and "comprehensive forecast light order demand signal".

6. The valet parking order prediction system based on fusion of multi-source data according to claim 5 is characterized by: The early warning feedback module is used to perform early warning control analysis and processing on the received relevant comprehensive order demand signals, generate order demand early warning signals of corresponding levels, and display and explain them. The specific process is as follows; When the "comprehensive prediction heavy order demand signal" is received, a first-level order demand warning signal is generated and sent to the display terminal in the form of a text message "It is predicted that the valet parking order in the current target area is in a heavy demand state", and a red flashing indicator light is lit; When the "comprehensive forecast medium order demand signal" is received, a secondary order demand warning signal is generated and sent to the display terminal in the form of a text message "It is predicted that the valet parking order in the current target area is in a medium demand state", and a yellow flashing indicator light is lit; When a "comprehensively predicted light order demand signal" is received, a third-level order demand warning signal is generated and sent to the display end for display in the form of text "It is predicted that the valet parking orders in the current target area are in a light demand state", and a blue flashing indicator light is turned on.

7. A valet parking order prediction method based on fusion of multi-source data, used to implement the valet parking order prediction system based on fusion of multi-source data as described in any one of claims 1 to 6, characterized in that: The steps include: Step 1: first, evenly divide the target area into several square areas, and connect the data source interface of the valet parking software through the application programming interface API to collect the service resource information of the parking service personnel in the target area, obtain the service resource information data, and collect the parking space resource information in the target area, obtain the parking space resource information data; Based on the analysis of the service resource information data and the parking space resource information data, a parking resource shortage index Zkf is obtained by fitting, and after comparative analysis, a relevant parking resource shortage signal is generated and sent to the order prediction module; Step 2: Secondly, divide the driving roads in the target area into several independent sections, monitor the traffic flow of several independent sections through multiple groups of traffic cameras, and obtain the traffic flow information data of each section after feature extraction; and analyze the congestion level in the target area to obtain the regional congestion index Zjt. After feature extraction and comparative analysis, generate relevant traffic congestion level signals and send them to the order prediction module; Step 3: Then, the received parking resource shortage signal and traffic congestion signal are processed for comprehensive order demand forecasting and analysis, and a comprehensive forecast order demand signal is generated, and sent to the early warning feedback module; Step 4: Finally, the received relevant comprehensive forecast order demand signals are analyzed and processed for early warning control, order demand early warning signals of corresponding levels are generated, and displayed and explained.

Images