Digital traffic law enforcement insight engine and research and judgment platform based on edge calculation

Abstract

The invention discloses a digital traffic law enforcement insight engine and research and judgment platform based on edge calculation, relates to the technical field of digital traffic law enforcement, and aims to solve the problem that the judgment and decision-making efficiency of transportation illegal behaviors is not high. Through information association and verification, the accuracy and integrity of data are ensured, reliable standard traffic research and judgment data are provided for a traffic and transportation comprehensive administrative law enforcement agency, more efficient decision making and execution are supported, the platform can automatically perform data extraction, fusion, association and verification, manual intervention is reduced, the management efficiency is improved, and the management cost is reduced. The platform reduces the workload of manual intervention and analysis through an automatic research and judgment process and intelligent analysis, improves the efficiency and accuracy of decision support, and further improves the processing efficiency and speed through GPU acceleration calculation of an AI chip, so that the system can meet the real-time processing requirements of large-scale data streams. And a data processing flow and an analysis model can be adjusted according to actual requirements.

Description

Technical Field

[0001] The present invention relates to the field of digital traffic law enforcement technology, and specifically to a digital traffic law enforcement insight engine and analysis platform based on edge computing. Background Art

[0002] Digital traffic enforcement refers to the entire process of using digital technology, information technology, and communication technology to supervise traffic behavior, detect violations, collect and process evidence, and execute and manage law enforcement activities.

[0003] The patent application with publication number CN118097967A discloses an intelligent management system for traffic law enforcement, which mainly divides roads into different law enforcement intervals through the calculation of traffic law enforcement coefficients by the law enforcement management module, and adopts different law enforcement methods according to different law enforcement intervals. When the traffic law enforcement coefficient belongs to interval law enforcement interval one, law enforcement can be carried out through traffic supervision equipment without the direct intervention of traffic law enforcement personnel, thereby saving law enforcement human resources; when the traffic law enforcement coefficient belongs to interval law enforcement interval two, law enforcement resources are flexibly deployed according to the situation of traffic facilities to improve law enforcement efficiency; so that law enforcement resources are optimally utilized, and law enforcement efficiency and coverage are improved. Although the above patent solves the problem of traffic supervision, the following problems still exist in actual operation: 1. The computing equipment and sensing equipment at the traffic monitoring points were not installed and data monitored more completely, resulting in inaccurate data collection at the traffic monitoring points.

[0004] 2. The real-time data of traffic monitoring points is not used to effectively predict and process abnormal events, resulting in reduced effectiveness in handling abnormal events.

[0005] 3. The actual situation of real-time traffic monitoring points is not used to make effective decisions and analyses of traffic abnormal events, resulting in reduced efficiency in handling abnormal events. Summary of the invention

[0006] The purpose of the present invention is to provide a digital traffic law enforcement insight engine and analysis platform based on edge computing, which ensures the accuracy and completeness of data through information association and verification, provides reliable standard traffic analysis data for comprehensive administrative law enforcement agencies of transportation, and supports more efficient decision-making and execution. The platform can automatically extract, fuse, associate and verify data, reduce manual intervention, improve management efficiency, and reduce the risk of human error. The platform reduces the burden of manual intervention and analysis through automated analysis processes and intelligent analysis, improves the efficiency and accuracy of decision support, and the GPU accelerated computing of AI chips further improves the processing efficiency and speed, enabling the system to cope with the real-time processing needs of large-scale data streams, and can adjust the data processing process and analysis model according to actual needs, which can solve the problems in the prior art.

[0007] To achieve the above object, the present invention provides the following technical solutions: The digital traffic law enforcement insight engine and analysis platform based on edge computing includes: Traffic data collection unit for: Deploy edge computing devices and sensing devices at traffic monitoring points. After deployment, different sensing devices will collect data. After data collection, traffic data to be analyzed will be obtained. Data fusion analysis unit, used for: The traffic data to be analyzed is fused. After the fusion is completed, edge computing technology is used to perform real-time data calculation. Edge computing technology is calculated through edge computing devices. After the real-time data calculation is completed, the traffic data to be predicted is obtained; Analytical data prediction and early warning unit, used for: The traffic data to be predicted will be used to predict abnormal events, and the degree of abnormality will be analyzed based on the predicted situation of the abnormal events. According to the analyzed degree of abnormality, an early warning will be issued at the traffic law enforcement terminal. Finally, the predicted abnormal events and the degree of abnormality of the abnormal events will be marked as traffic abnormality data.

[0008] Preferably, the traffic data collection unit is also used for: First, confirm the edge computing equipment and perception equipment. The edge computing equipment includes circuit boards, main control devices, AI chips, storage devices, interface devices, and temperature control devices; the perception equipment includes signal equipment, camera equipment, and radar equipment; Install the circuit board, main control device, AI chip, storage device, interface device and temperature control device in the designed chassis, fix the chassis in a predetermined position, and connect the power supply and network; the signal device is installed on the signal light pole of the traffic monitoring point, the camera device is installed in a position that can cover the entire traffic monitoring point, and the radar device is installed in a place that can detect the speed and position of the vehicle. At the same time, the radar device and the camera device are used in conjunction with each other; When the sensing device collects data from the traffic monitoring points, each sensing device is first tested, and data collection is performed after all the devices are qualified. The traffic monitoring points include highway overload detection stations, dynamic monitoring points, expressway toll stations, freight sources, provincial boundary non-stop gantries, ferry crossings, comprehensive law enforcement workstations, etc. The signal equipment collects data on signal phase, signal timing and traffic violations; the camera equipment collects data on vehicle images, vehicle video streams, vehicle trajectories, traffic flow and average speed; the radar equipment collects data on vehicle speed, vehicle spacing and vehicle position; After all the data are collected, the traffic data to be analyzed are obtained.

[0009] Preferably, the radar device is used in conjunction with the camera device, including: The radar device collects vehicle speed in real time; When the vehicle speed exceeds a preset first vehicle speed threshold, the camera is triggered to collect vehicle tracking video streams for the vehicle; When the vehicle speed does not exceed a preset first vehicle speed threshold, but exceeds a preset second vehicle speed threshold, a vehicle corresponding to the vehicle whose speed exceeds the preset second vehicle speed threshold is taken as a target vehicle, and the target vehicle is marked; Retrieving the vehicle driving parameters of the target vehicle obtained by the radar equipment at each traffic monitoring point previously passed by the target vehicle, wherein the vehicle driving parameters include the vehicle speed of the target vehicle, the vehicle distance between the target vehicle and the rear vehicle, the vehicle distance between the target vehicle and the front vehicle, and the vehicle speeds of the rear vehicle and the front vehicle; Using the vehicle driving parameters of the target vehicle at each traffic monitoring point, obtaining a first vehicle driving safety factor of the target vehicle at each traffic monitoring point; The first vehicle driving safety factor of the target vehicle at each traffic monitoring point is obtained by the following formula: ; Among them, S 01 represents the first vehicle driving safety factor of the target vehicle at each traffic monitoring point; D x and D h V represents the vehicle distance between the target vehicle and the preceding vehicle and the vehicle distance between the target vehicle and the following vehicle respectively; x and V h are the vehicle speeds of the front vehicle and the rear vehicle respectively; V d Indicates the vehicle speed of the target vehicle; D a Indicates the preset distance reference value; V c represents a preset vehicle speed reference value; A represents a reference constant, and the value range of the reference constant is 1.5-100; Comparing the first vehicle driving safety factor of each traffic monitoring point with a preset safety factor threshold; When the first vehicle driving safety factor of the target vehicle at any traffic monitoring point is lower than a preset safety factor threshold, the camera is triggered to collect vehicle tracking video streams for the target vehicle; When the first vehicle driving safety factor of the target vehicle at each traffic monitoring point is not lower than the preset safety factor threshold, the safety of the target vehicle is continuously judged, and when the safety of the target vehicle is lower than the safety requirement, the camera is triggered to perform vehicle tracking video stream acquisition on the target vehicle.

[0010] Preferably, when the first vehicle driving safety factor of the target vehicle at each traffic monitoring point is not lower than a preset safety factor threshold, the safety of the target vehicle is continuously determined, and when the safety of the target vehicle is lower than the safety requirement, the camera is triggered to perform vehicle tracking video stream acquisition on the target vehicle, including: When the first vehicle driving safety factor of the target vehicle at each traffic monitoring point is not lower than a preset safety factor threshold, extracting the first vehicle driving safety factor of the target vehicle at each traffic monitoring point; Obtaining a comprehensive vehicle driving safety factor using the first vehicle driving safety factor of the target vehicle at each traffic monitoring point; Extracting vehicle travel parameters of each subsequent traffic monitoring point after the vehicle speed of the target vehicle exceeds a preset second vehicle speed threshold, wherein the vehicle travel parameters include the vehicle speed of the target vehicle, the vehicle distance between the target vehicle and a rear vehicle, the vehicle distance between the target vehicle and a front vehicle, and the vehicle speeds of the rear vehicle and the front vehicle; Obtaining a second vehicle driving safety factor corresponding to the target vehicle at each subsequent traffic monitoring point by using the vehicle driving parameter of each subsequent traffic monitoring point after the vehicle speed of the target vehicle exceeds a preset second vehicle speed threshold; The second vehicle driving safety factor corresponding to each subsequent traffic monitoring point of the target vehicle is obtained by the following formula: ; Among them, S 02 represents the second vehicle driving safety factor corresponding to each subsequent traffic monitoring point of the target vehicle; D mx and D mh V respectively represent the vehicle distance between the target vehicle and the preceding vehicle and the vehicle distance between the target vehicle and the following vehicle corresponding to each subsequent traffic monitoring point; V mx and V mh Respectively represent the vehicle speed of the preceding vehicle and the vehicle speed of the following vehicle corresponding to the target vehicle corresponding to each subsequent traffic monitoring point; V d Indicates the vehicle speed of the target vehicle; D a represents a preset distance reference value; A represents a reference constant, and the value range of the reference constant is 1.5-100; using the comprehensive vehicle driving safety factor to compare with a second vehicle driving safety factor corresponding to the target vehicle at each subsequent traffic monitoring point; When the second vehicle driving safety factor corresponding to the target vehicle at each subsequent traffic monitoring point is lower than the comprehensive vehicle driving safety factor, the camera is triggered to perform vehicle tracking video stream acquisition on the target vehicle.

[0011] Preferably, obtaining a comprehensive vehicle driving safety factor by using the first vehicle driving safety factor of the target vehicle at each traffic monitoring point includes: Extract the vehicle flow and vehicle speed corresponding to each unit time of each traffic monitoring point; wherein the value range of the unit time is 10s-30s; Obtain a weight value corresponding to each traffic monitoring point using the vehicle flow and vehicle speed corresponding to each unit time corresponding to each traffic monitoring point; The weight value corresponding to each traffic monitoring point is obtained by the following formula: ; Among them, W j represents the weight value corresponding to the jth traffic monitoring point; n represents the number of vehicles running per unit time at each traffic monitoring point; C i V represents the traffic volume per i-th unit time corresponding to each traffic monitoring point; i represents the average vehicle speed per unit time corresponding to each traffic monitoring point; T i represents the traffic light control response delay rate of the i-th unit time corresponding to each traffic monitoring point; Obtaining a comprehensive vehicle driving safety factor by using the weight value corresponding to each traffic monitoring point and the first vehicle driving safety factor corresponding to each traffic monitoring point; The comprehensive vehicle driving safety factor is obtained by the following formula: ; Among them, S c represents the comprehensive vehicle driving safety factor; m represents the number of traffic monitoring points; W j represents the weight value corresponding to the jth traffic monitoring point; C pj represents the average traffic flow corresponding to the jth traffic monitoring point; S 01j Represents the first vehicle driving safety factor corresponding to the j-th traffic monitoring point.

[0012] Preferably, the data fusion analysis unit is further used for: The collected traffic data to be analyzed is preprocessed. The data preprocessing process includes data cleaning, data alignment, data conversion and data standardization; After data preprocessing is completed, time synchronization and space synchronization are performed. Time synchronization includes time standard unification, time alignment, missing time processing and time window division. Space synchronization includes coordinate system one, space alignment, missing space processing and space division. After the time synchronization and space synchronization are processed respectively, the time-space index is established and verified, and finally the traffic data to be analyzed after the time synchronization and space synchronization are completed are obtained, and the features of the traffic data to be analyzed after the time synchronization and space synchronization are completed are extracted; After feature extraction is completed, the feature data collected by different devices are associated and matched; The sensor fusion algorithm is used to fuse the associated and matched data.

[0013] Preferably, the data fusion analysis unit is further used for: The traffic data to be analyzed after data fusion is input into the edge computing device, and the edge computing device uses streaming data processing technology to process the input real-time data stream; The process of processing the input real-time data stream includes: According to the deep learning model deployed on the edge computing device, the input real-time data stream is inferred in real time, and AI inference data is obtained after real-time inference; Use GPU accelerated computing of AI chips to perform parallel and distributed computing on AI inference data; After parallel and distributed computing is completed, the traffic data to be predicted is obtained.

[0014] Preferably, the analysis data prediction and early warning unit is also used for: Obtain a random forest model and historical traffic data from a database, use the historical traffic data to train the random forest model, and obtain a standard random forest model after the training is completed; Extracting features for abnormal event prediction from the data to be predicted, the features including traffic flow change rate, vehicle speed abnormal value and signal phase duration; The extracted features are input into the standard random forest model to infer abnormal events. Before inferring abnormal events, the output inferred abnormal events are confirmed by setting a threshold. According to the abnormal threshold range of the inferred abnormal event, the abnormal degree of the abnormal event is confirmed, and the abnormal degree is divided into level one abnormality, level two abnormality and level three abnormality; Then, the warning level at the traffic law enforcement terminal is confirmed according to the degree of abnormality, and the warning levels are divided into level one, level two and level three. Finally, the inferred abnormal events and the abnormal degree of the abnormal events are uniformly labeled as traffic anomaly data.

[0015] The digital traffic law enforcement analysis platform based on edge computing includes: Data integration unit for: Integrate traffic anomaly data with analysis information, and obtain standard traffic analysis data after data integration; Among them, the abnormal data in the traffic abnormal data is extracted, and the abnormal data includes the abnormal event type, abnormal degree, occurrence time and event location; Confirm the assessment information, which includes weather information and special time information; The traffic anomaly data is integrated with the judgment information of the same time and place. At the same time, the characteristics of traffic anomalies are combined with the characteristics of weather and special time information to form a comprehensive feature set; According to traffic management and law enforcement rules, establish the association rules between traffic anomalies and analysis information; Associating association rules with comprehensive feature sets, and verifying accuracy and completeness after associating information; After accuracy and completeness verification is completed, standard traffic analysis data is obtained.

[0016] Preferably, it also includes: Integrated data analysis and decision-making unit, used for: The standard traffic analysis data is analyzed and judged using historical violation data and professional data, and finally the analysis and judgment results are transmitted to different processing centers; Extract historical violation data and professional data from the database. Historical violation data includes violations, penalty records and accident situations; professional data includes traffic rules, laws and regulations, traffic flow statistics and historical accident analysis; Formulate research and judgment rules based on professional data, including the association rules between abnormal events and illegal behaviors, and convert the formulated research and judgment rules into operational parameters; Conduct similar trend analysis on key feature data in standard traffic analysis data based on historical violation data. Key feature data include abnormal event type, abnormal degree, occurrence time and location; Use professional data to provide regulatory basis and historical case comparison for standard traffic analysis data completed for similar trend analysis; After the legal basis is provided and compared with historical cases, the traffic analysis data to be judged is obtained; Input the traffic analysis data to be judged into the analysis model for real-time analysis and judgment; Real-time analysis and judgment is to judge the traffic analysis data according to the analysis rules and make law enforcement action judgments; After real-time analysis and judgment of the traffic analysis data to be judged, the real-time traffic analysis result is obtained; Finally, the real-time traffic analysis results are transmitted to the traffic command center and law enforcement agencies.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. The digital traffic law enforcement insight engine and analysis platform based on edge computing provided by the present invention, the edge computing equipment can dynamically allocate computing resources according to actual needs, avoid excessive concentration and waste of resources, improve resource utilization efficiency, and the precise deployment and coordinated use of sensing equipment can realize comprehensive and high-precision data collection of traffic monitoring points, providing a reliable data foundation for traffic law enforcement and analysis.

[0018] 2. The digital traffic law enforcement insight engine and analysis platform based on edge computing provided by the present invention and the GPU accelerated computing of the AI ​​chip further improve the processing efficiency and speed, enabling the system to cope with the real-time processing needs of large-scale data streams. The data processing flow and analysis model can be adjusted according to actual needs to adapt to different traffic law enforcement scenarios and needs. Through the rapid inference capability of the model, it can monitor and respond to abnormal traffic events in real time, and enhance the dynamic management capabilities of urban traffic systems.

[0019] 3. The digital traffic law enforcement insight engine and analysis platform based on edge computing provided by the present invention ensures the accuracy and completeness of data through information association and verification, provides reliable standard traffic analysis data for comprehensive administrative law enforcement agencies of transportation, and supports more efficient decision-making and execution. The platform can automatically extract, fuse, associate and verify data, reduce manual intervention, improve management efficiency, and reduce the risk of human error. The platform reduces the burden of manual intervention and analysis through automated analysis processes and intelligent analysis, and improves the efficiency and accuracy of decision support. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A schematic diagram of a unit of a digital traffic law enforcement insight engine and analysis platform of the present invention; Figure 2 It is a flow chart of the digital traffic law enforcement insight engine and analysis platform of the present invention. DETAILED DESCRIPTION

[0021] 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.

[0022] In order to solve the problem in the prior art that the computing equipment and sensing equipment at the traffic monitoring point are not installed and monitored more completely, resulting in inaccurate data collection at the traffic monitoring point, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions: The digital traffic law enforcement insight engine and analysis platform based on edge computing includes: Traffic data collection unit for: Deploy edge computing devices and sensing devices at traffic monitoring points. After deployment, different sensing devices will collect data. After data collection, traffic data to be analyzed will be obtained. Data fusion analysis unit, used for: The traffic data to be analyzed is fused. After the fusion is completed, edge computing technology is used to perform real-time data calculation. Edge computing technology is calculated through edge computing devices. After the real-time data calculation is completed, the traffic data to be predicted is obtained; Analytical data prediction and early warning unit, used for: The traffic data to be predicted will be used to predict abnormal events, and the degree of abnormality will be analyzed based on the predicted situation of the abnormal events. According to the analyzed degree of abnormality, an early warning will be issued at the traffic law enforcement terminal. Finally, the predicted abnormal events and the degree of abnormality of the abnormal events will be marked as traffic abnormality data.

[0023] Specifically, the edge computing device of the traffic data collection unit can quickly analyze and respond to traffic data, which helps to realize instant traffic management decisions and law enforcement actions, and improves the response speed and safety of the traffic system. The edge computing device of the data fusion analysis unit can process the input real-time data stream in real time, reducing the delay and bandwidth requirements of data transmission. The data prediction and early warning unit can monitor and respond to abnormal traffic events in real time through the rapid inference capability of the model through analysis, thereby improving the dynamic management capabilities of the urban traffic system.

[0024] Traffic data collection unit is also used for: First, confirm the edge computing equipment and perception equipment. The edge computing equipment includes circuit boards, main control devices, AI chips, storage devices, interface devices, and temperature control devices; the perception equipment includes signal equipment, camera equipment, and radar equipment; Install the circuit board, main control device, AI chip, storage device, interface device and temperature control device in the designed chassis, fix the chassis in a predetermined position, and connect the power supply and network; the signal device is installed on the signal light pole of the traffic monitoring point, the camera device is installed in a position that can cover the entire traffic monitoring point, and the radar device is installed in a place that can detect the speed and position of the vehicle. At the same time, the radar device and the camera device are used in conjunction with each other; When the sensing device collects data from the traffic monitoring points, each sensing device is first tested, and data collection is performed after all the devices are qualified. The traffic monitoring points include highway overload detection stations, dynamic monitoring points, expressway toll stations, freight sources, provincial boundary non-stop gantries, ferry crossings, comprehensive law enforcement workstations, etc. The signal equipment collects data on signal phase, signal timing and traffic violations; the camera equipment collects data on vehicle images, vehicle video streams, vehicle trajectories, traffic flow and average speed; the radar equipment collects data on vehicle speed, vehicle spacing and vehicle position; After all the data are collected, the traffic data to be analyzed are obtained.

[0025] Specifically, by deploying edge computing devices near traffic monitoring points, data from sensing devices can be processed in real time or near real time, reducing the time and bandwidth required for data transmission to remote data centers, thereby improving data processing efficiency. Edge computing devices can quickly analyze and respond to traffic data, which helps to achieve instant traffic management decisions and law enforcement actions, and improve the response speed and security of the transportation system. Data is initially processed and analyzed at the edge level, which reduces the transmission volume of sensitive data, reduces the risk of data leakage, and enhances data security. Edge computing devices can dynamically allocate computing resources according to actual needs, avoid excessive concentration and waste of resources, and improve resource utilization efficiency. The precise deployment and coordinated use of sensing equipment (such as signal machines, cameras, and radars) can achieve comprehensive and high-precision data collection of traffic monitoring points, providing a reliable data basis for traffic law enforcement and analysis. By collecting and analyzing traffic data in real time, traffic congestion, violations, and other problems can be discovered in a timely manner, and corresponding measures can be taken to intervene and manage, thereby improving the operating efficiency and safety of the transportation system.

[0026] In order to solve the problem that the existing technology does not effectively predict and process abnormal events based on the real-time data of traffic monitoring points, which leads to reduced abnormal event processing effect, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions: Specifically, the radar device is used in conjunction with the camera device, including: The radar device collects vehicle speed in real time; When the vehicle speed exceeds a preset first vehicle speed threshold, the camera is triggered to collect vehicle tracking video streams for the vehicle; When the vehicle speed does not exceed a preset first vehicle speed threshold, but exceeds a preset second vehicle speed threshold, a vehicle corresponding to the vehicle whose speed exceeds the preset second vehicle speed threshold is taken as a target vehicle, and the target vehicle is marked; Retrieving the vehicle driving parameters of the target vehicle obtained by the radar equipment at each traffic monitoring point previously passed by the target vehicle, wherein the vehicle driving parameters include the vehicle speed of the target vehicle, the vehicle distance between the target vehicle and the rear vehicle, the vehicle distance between the target vehicle and the front vehicle, and the vehicle speeds of the rear vehicle and the front vehicle; Using the vehicle driving parameters of the target vehicle at each traffic monitoring point, obtaining a first vehicle driving safety factor of the target vehicle at each traffic monitoring point; The first vehicle driving safety factor of the target vehicle at each traffic monitoring point is obtained by the following formula: ; Among them, S 01 represents the first vehicle driving safety factor of the target vehicle at each traffic monitoring point; D x and D h V represents the vehicle distance between the target vehicle and the preceding vehicle and the vehicle distance between the target vehicle and the following vehicle respectively; x and V h are the vehicle speeds of the front vehicle and the rear vehicle respectively; V d Indicates the vehicle speed of the target vehicle; D a Indicates the preset distance reference value; V c represents a preset vehicle speed reference value; A represents a reference constant, and the value range of the reference constant is 1.5-100; Comparing the first vehicle driving safety factor of each traffic monitoring point with a preset safety factor threshold; When the first vehicle driving safety factor of the target vehicle at any traffic monitoring point is lower than a preset safety factor threshold, the camera is triggered to collect vehicle tracking video streams for the target vehicle; When the first vehicle driving safety factor of the target vehicle at each traffic monitoring point is not lower than the preset safety factor threshold, the safety of the target vehicle is continuously judged, and when the safety of the target vehicle is lower than the safety requirement, the camera is triggered to perform vehicle tracking video stream acquisition on the target vehicle.

[0027] The technical effect of the above technical solution is: by collecting vehicle speed in real time through radar equipment, it can respond quickly to changes in traffic conditions. When the vehicle speed exceeds the preset threshold, the camera can immediately trigger the vehicle tracking video stream acquisition, which helps to capture and deal with potential dangerous situations in a timely manner. By using the collaborative work of radar and camera, speeding vehicles can be accurately identified and located, and target vehicles with speeds between two thresholds can be marked. This helps traffic management departments to classify and manage vehicles more carefully and improve law enforcement efficiency and accuracy. By retrieving the driving parameters of the target vehicle at different traffic monitoring points, including speed, vehicle spacing, etc., the driving safety of the vehicle can be comprehensively evaluated. Using these parameters to calculate the first vehicle driving safety factor at each traffic monitoring point can more objectively reflect the driving status and safety performance of the vehicle. When the first vehicle driving safety factor at any traffic monitoring point is lower than the preset safety factor threshold, the camera will automatically trigger the vehicle tracking video stream acquisition, which helps to timely discover and deal with potential safety hazards. Even if the safety factor is not lower than the threshold, the system will continue to judge the safety of the target vehicle and trigger an early warning when the safety is lower than the requirement, which reflects the intelligence and foresight of the system. Through the automated monitoring and early warning system, the subjectivity of manual intervention and judgment can be reduced, and the objectivity and fairness of traffic management can be improved. At the same time, the real-time and response speed of the system also help traffic management departments to respond to emergencies such as traffic congestion and accidents more quickly, and improve the overall traffic operation efficiency. By real-time monitoring and analyzing the driving status of vehicles, the system can promptly detect and correct the driver's violations, such as speeding and not maintaining a safe distance, thereby effectively reducing the occurrence of traffic accidents.

[0028] In summary, the technical effects of this technical solution in terms of performance indicators are mainly reflected in real-time and response speed, accurate identification and positioning, multi-dimensional data analysis, intelligent warning and intervention, improved traffic management efficiency and enhanced road safety. These effects jointly improve the efficiency and accuracy of traffic monitoring and vehicle safety management, and provide strong support for building a safer and more efficient traffic environment.

[0029] Specifically, when the first vehicle driving safety factor of the target vehicle at each traffic monitoring point is not lower than the preset safety factor threshold, the safety of the target vehicle is continuously determined, and when the safety of the target vehicle is lower than the safety requirement, the camera is triggered to perform vehicle tracking video stream acquisition on the target vehicle, including: When the first vehicle driving safety factor of the target vehicle at each traffic monitoring point is not lower than a preset safety factor threshold, extracting the first vehicle driving safety factor of the target vehicle at each traffic monitoring point; Obtaining a comprehensive vehicle driving safety factor using the first vehicle driving safety factor of the target vehicle at each traffic monitoring point; Extracting vehicle travel parameters of each subsequent traffic monitoring point after the vehicle speed of the target vehicle exceeds a preset second vehicle speed threshold, wherein the vehicle travel parameters include the vehicle speed of the target vehicle, the vehicle distance between the target vehicle and a rear vehicle, the vehicle distance between the target vehicle and a front vehicle, and the vehicle speeds of the rear vehicle and the front vehicle; Obtaining a second vehicle driving safety factor corresponding to the target vehicle at each subsequent traffic monitoring point by using the vehicle driving parameter of each subsequent traffic monitoring point after the vehicle speed of the target vehicle exceeds a preset second vehicle speed threshold; The second vehicle driving safety factor corresponding to each subsequent traffic monitoring point of the target vehicle is obtained by the following formula: ; Among them, S 02 represents the second vehicle driving safety factor corresponding to each subsequent traffic monitoring point of the target vehicle; D mx and D mh V respectively represent the vehicle distance between the target vehicle and the preceding vehicle and the vehicle distance between the target vehicle and the following vehicle corresponding to each subsequent traffic monitoring point; V mx and V mh Respectively represent the vehicle speed of the preceding vehicle and the vehicle speed of the following vehicle corresponding to the target vehicle corresponding to each subsequent traffic monitoring point; V d Indicates the vehicle speed of the target vehicle; D a represents a preset distance reference value; A represents a reference constant, and the value range of the reference constant is 1.5-100; using the comprehensive vehicle driving safety factor to compare with a second vehicle driving safety factor corresponding to the target vehicle at each subsequent traffic monitoring point; When the second vehicle driving safety factor corresponding to the target vehicle at each subsequent traffic monitoring point is lower than the comprehensive vehicle driving safety factor, the camera is triggered to perform vehicle tracking video stream acquisition on the target vehicle.

[0030] The technical effect of the above technical solution is: by extracting the first vehicle driving safety factor of the target vehicle at each traffic monitoring point and calculating the comprehensive vehicle driving safety factor, the overall driving safety of the target vehicle can be more comprehensively evaluated. By using the vehicle driving parameters of each subsequent traffic monitoring point after the target vehicle exceeds the second vehicle speed threshold, the second vehicle driving safety factor is calculated, which can accurately capture the changes in the driving state of the target vehicle and improve the accuracy and timeliness of the judgment. When the second vehicle driving safety factor is lower than the comprehensive vehicle driving safety factor, the system can immediately trigger the camera to perform vehicle tracking video stream acquisition on the target vehicle. This setting enables the system to issue an early warning in time when the driving safety of the target vehicle shows a downward trend, providing the traffic management department with a timely intervention opportunity to reduce potential safety risks. When the first vehicle driving safety factor of each traffic monitoring point is not lower than the preset safety factor threshold, the system will not immediately trigger the camera for acquisition, but will continue to judge. This design avoids unnecessary waste of resources, so that camera resources can be used more efficiently to handle vehicles that really need attention. Through automated monitoring and judgment processes, the system can quickly identify and handle potential safety hazards and improve the efficiency and accuracy of traffic management. At the same time, the system can also provide real-time vehicle driving data and safety assessment results to traffic management departments, helping management departments make more scientific and reasonable decisions. By real-time monitoring and evaluating the driving safety of vehicles, the system can promptly detect and correct drivers' violations, such as speeding and failure to maintain a safe distance. This setting helps to improve drivers' safety awareness and the consciousness of abiding by traffic rules, thereby further enhancing road safety.

[0031] In summary, the technical effects of this technical solution in terms of performance indicators are mainly reflected in enhancing judgment accuracy, improving warning sensitivity, optimizing resource utilization, improving traffic management efficiency, and enhancing road safety. These effects have jointly improved the intelligent level of traffic monitoring and vehicle safety management, and provided strong support for building a safer and more efficient traffic environment.

[0032] Specifically, using the first vehicle driving safety factor of the target vehicle at each traffic monitoring point to obtain a comprehensive vehicle driving safety factor includes: Extract the vehicle flow and vehicle speed corresponding to each unit time of each traffic monitoring point; wherein the value range of the unit time is 10s-30s; Obtain a weight value corresponding to each traffic monitoring point using the vehicle flow and vehicle speed corresponding to each unit time corresponding to each traffic monitoring point; The weight value corresponding to each traffic monitoring point is obtained by the following formula: ; Among them, W j represents the weight value corresponding to the jth traffic monitoring point; n represents the number of vehicles running per unit time at each traffic monitoring point; C i V represents the traffic volume per i-th unit time corresponding to each traffic monitoring point; i represents the average vehicle speed per unit time corresponding to each traffic monitoring point; T i represents the traffic light control response delay rate of the i-th unit time corresponding to each traffic monitoring point; Obtaining a comprehensive vehicle driving safety factor by using the weight value corresponding to each traffic monitoring point and the first vehicle driving safety factor corresponding to each traffic monitoring point; The comprehensive vehicle driving safety factor is obtained by the following formula: ; Among them, S c represents the comprehensive vehicle driving safety factor; m represents the number of traffic monitoring points; W j represents the weight value corresponding to the jth traffic monitoring point; C pj represents the average traffic flow corresponding to the jth traffic monitoring point; S 01j Represents the first vehicle driving safety factor corresponding to the j-th traffic monitoring point.

[0033] The technical effect of the above technical solution is: by considering the traffic volume and vehicle speed of each traffic monitoring point in different unit time, the technical solution can more comprehensively evaluate the driving safety of the target vehicle under different traffic conditions. The introduction of the traffic light control response delay rate as a factor in the weight calculation further improves the accuracy and practicality of the evaluation. The weight value of each traffic monitoring point is calculated by a formula to ensure the objectivity and scientificity of the weight distribution. The introduction of the weight value enables different traffic monitoring points to obtain different weights according to their importance in the calculation of the comprehensive vehicle driving safety factor, which improves the accuracy and pertinence of the evaluation. By combining the weight value of each traffic monitoring point and the first vehicle driving safety factor, the comprehensive vehicle driving safety factor is calculated, which can more accurately reflect the driving safety of the target vehicle at different traffic monitoring points. This comprehensive evaluation method helps the traffic management department to have a more comprehensive understanding of the driving status of the target vehicle, so as to make more scientific decisions. When the comprehensive vehicle driving safety factor is lower than the preset safety threshold, the system can issue an early warning in time to remind the traffic management department to pay attention and take corresponding measures. This early warning mechanism helps to timely discover and deal with potential safety hazards and improve the efficiency and safety of traffic management. By comprehensively considering the weights and driving safety factors of different traffic monitoring points, this technical solution can allocate monitoring resources more reasonably and avoid unnecessary waste. This way of optimizing resource utilization helps to improve the overall performance and efficiency of the traffic monitoring system. Through an automated evaluation process, this technical solution can quickly and accurately provide driving safety information of target vehicles and provide decision support for traffic management departments. This efficient management method helps to improve the overall efficiency of traffic management and reduce traffic congestion and accidents.

[0034] In summary, the technical effects of this technical solution in terms of performance indicators are mainly reflected in improving the comprehensiveness of the evaluation, optimizing the weight distribution, improving the accuracy of the evaluation, enhancing the early warning capability, optimizing resource utilization, and improving the efficiency of traffic management. These effects have jointly improved the intelligence level and practical performance of the traffic monitoring system, and provided strong support for building a safer and more efficient traffic environment.

[0035] The data fusion analysis unit is also used for: The collected traffic data to be analyzed is preprocessed. The data preprocessing process includes data cleaning, data alignment, data conversion and data standardization; After data preprocessing is completed, time synchronization and space synchronization are performed. Time synchronization includes time standard unification, time alignment, missing time processing and time window division. Space synchronization includes coordinate system one, space alignment, missing space processing and space division. After the time synchronization and space synchronization are processed respectively, the time-space index is established and verified, and finally the traffic data to be analyzed after the time synchronization and space synchronization are completed are obtained, and the features of the traffic data to be analyzed after the time synchronization and space synchronization are completed are extracted; After feature extraction is completed, the feature data collected by different devices are associated and matched; The sensor fusion algorithm is used to fuse the associated and matched data.

[0036] The traffic data to be analyzed after data fusion is input into the edge computing device, and the edge computing device uses streaming data processing technology to process the input real-time data stream; The process of processing the input real-time data stream includes: According to the deep learning model deployed on the edge computing device, the input real-time data stream is inferred in real time, and AI inference data is obtained after real-time inference; Use GPU accelerated computing of AI chips to perform parallel and distributed computing on AI inference data; After parallel and distributed computing is completed, the traffic data to be predicted is obtained.

[0037] Specifically, data cleaning, alignment, conversion, and standardization steps ensure the quality and consistency of input data, providing a solid foundation for subsequent analysis and processing. The preprocessing step reduces noise and outliers, improving the accuracy and reliability of data. Time synchronization ensures the consistency and coherence of data in time, facilitating time series analysis and historical comparison. Spatial synchronization enables data from different sources to be compared and analyzed in the same geographic reference system, improving the accuracy of spatial analysis. The establishment of spatiotemporal indexes improves the efficiency of data retrieval and query, making it possible to quickly locate and analyze data within a specific spatiotemporal range. Feature extraction can mine useful features from complex data. information, providing key indicators for subsequent analysis. Edge computing devices can process input real-time data streams in real time, reducing data transmission latency and bandwidth requirements. The real-time inference capability of deep learning models enables the system to quickly identify and judge traffic conditions, providing strong support for real-time traffic law enforcement. The GPU accelerated computing of AI chips further improves processing efficiency and speed, enabling the system to cope with real-time processing needs of large-scale data streams. The data processing process and analysis model can be adjusted according to actual needs to adapt to different traffic law enforcement scenarios and needs. By real-time processing and analysis of traffic data, the system can provide law enforcement personnel with accurate and timely traffic status information.

[0038] The data analysis, prediction and early warning unit is also used to: Obtain a random forest model and historical traffic data from a database, use the historical traffic data to train the random forest model, and obtain a standard random forest model after the training is completed; Extracting features for abnormal event prediction from the data to be predicted, the features including traffic flow change rate, vehicle speed abnormal value and signal phase duration; The extracted features are input into the standard random forest model to infer abnormal events. Before inferring abnormal events, the output inferred abnormal events are confirmed by setting a threshold. According to the abnormal threshold range of the inferred abnormal event, the abnormal degree of the abnormal event is confirmed, and the abnormal degree is divided into level one abnormality, level two abnormality and level three abnormality; Then, the warning level at the traffic law enforcement terminal is confirmed according to the degree of abnormality, and the warning levels are divided into level one, level two and level three. Finally, the inferred abnormal events and the abnormal degree of the abnormal events are uniformly labeled as traffic anomaly data.

[0039] Specifically, historical traffic data is used to train the random forest model, so that the model can make accurate predictions based on a large amount of historical information, which improves the reliability and accuracy of the model. By extracting multi-dimensional features such as traffic flow change rate, vehicle speed anomalies, and signal phase duration, the model can more comprehensively analyze traffic conditions and identify abnormal events. The random forest model has good adaptability and can handle complex and nonlinear traffic data, and maintain stable performance when feature dimensions increase or change. By setting thresholds and dividing the degree of abnormality, the solution can accurately identify and classify abnormal events, which helps to take corresponding measures for events of different severity and optimize resource allocation. The division of warning levels enables traffic management departments to respond quickly according to the severity of the event, improving the efficiency and timeliness of traffic management. Since a machine learning model is used, the solution can continuously update the model. and add new features to adapt to changes in traffic patterns and the emergence of new problems. Through the rapid inference capability of the model, the solution can monitor and respond to abnormal traffic events in real time, enhance the dynamic management capability of the urban transportation system, and edge computing can autonomously process data locally and send the final processing results to the cloud computing platform, reducing the amount of data transmitted over the network, thereby reducing the burden on the network. Among them, the warning types include daily warnings and interception warnings. Daily warnings include overdue annual inspections of operating vehicles, expired operating vehicle certificates, expected vehicle technical grade assessments, long-term offline warnings for "two passengers and one dangerous goods" vehicles, and long-term off-site operations of dangerous goods transport vehicles from other provinces; interception warnings include warnings for key operating vehicles that are offline for more than 24 hours due to satellite positioning, illegal operations of heavy-loaded general cargo vehicles, illegal operations of dangerous goods transport vehicles during prohibited operating periods, warnings for the assessment of overloading and speeding violations, and electronic fence warnings.

[0040] In order to solve the problem in the prior art that the actual situation of real-time traffic monitoring points is not used to make effective decisions and judgments on traffic abnormal events, thus reducing the efficiency of abnormal event processing, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions: The digital traffic law enforcement insight engine and analysis platform based on edge computing also includes: Data integration unit for: Integrate traffic anomaly data with analysis information, and obtain standard traffic analysis data after data integration; Integrated data analysis and decision-making unit, used for: Standard traffic analysis data is analyzed and judged using historical violation data and professional data, and the results are finally transmitted to different processing centers.

[0041] Specifically, by integrating traffic anomaly data with weather and special time information through the data integration unit to form a comprehensive feature set, the background and causes of traffic anomalies can be more comprehensively understood and analyzed. By integrating the trend analysis of historical violation data in the data analysis and decision-making unit and comparing the legal basis and historical cases, it provides important background and basis for the analysis and judgment, ensuring the reliability and accuracy of the analysis and judgment.

[0042] The data integration unit is also used for: Extracting abnormal data from traffic abnormal data, the abnormal data includes abnormal event type, abnormal degree, occurrence time and event location; Confirm the assessment information, which includes weather information and special time information; The traffic anomaly data is integrated with the judgment information of the same time and place. At the same time, the characteristics of traffic anomalies are combined with the characteristics of weather and special time information to form a comprehensive feature set; According to traffic management and law enforcement rules, establish the association rules between traffic anomalies and analysis information; Associating association rules with comprehensive feature sets, and verifying accuracy and completeness after associating information; After accuracy and completeness verification is completed, standard traffic analysis data is obtained.

[0043] Specifically, through edge computing technology, data processing is performed at a location close to the data generation source, which can realize real-time extraction and analysis of traffic anomalies and assessment information, reduce latency, and improve response speed. By integrating traffic anomaly data with weather and special time information to form a comprehensive feature set, the background and causes of traffic anomalies can be more comprehensively understood and analyzed, and the accuracy of abnormal event identification can be improved. Establishing association rules between traffic anomalies and assessment information will help reveal potential patterns and causal relationships, and provide more reliable decision-making support for traffic management and law enforcement. Through information association and verification, the accuracy and completeness of data are ensured, and reliable standard traffic assessment data is provided to comprehensive administrative law enforcement agencies for transportation, supporting more efficient decision-making and execution. The platform can automatically extract, fuse, associate and verify data, reduce manual intervention, improve management efficiency, and reduce the risk of human error. It can be adjusted and expanded according to changes in traffic management needs, support new data sources and analysis dimensions, and maintain the flexibility and adaptability of the system. Through accurate traffic anomaly identification and assessment, it helps traffic law enforcement departments respond and handle abnormal events more quickly, improve law enforcement efficiency and traffic safety.

[0044] The integrated data analysis and decision-making unit is also used for: Extract historical violation data and professional data from the database. Historical violation data includes violations, penalty records and accident situations; professional data includes traffic rules, laws and regulations, traffic flow statistics and historical accident analysis; Formulate research and judgment rules based on professional data, including the association rules between abnormal events and illegal behaviors, and convert the formulated research and judgment rules into operational parameters; Conduct similar trend analysis on key feature data in standard traffic analysis data based on historical violation data. Key feature data include abnormal event type, abnormal degree, occurrence time and location; Use professional data to provide regulatory basis and historical case comparison for standard traffic analysis data completed for similar trend analysis; After the legal basis is provided and compared with historical cases, the traffic analysis data to be judged is obtained; Input the traffic analysis data to be judged into the analysis model for real-time analysis and judgment; Real-time analysis and judgment is to judge the traffic analysis data according to the analysis rules and make law enforcement action judgments; After real-time analysis and judgment of the traffic analysis data to be judged, the real-time traffic analysis result is obtained; Finally, the real-time traffic analysis results are transmitted to the traffic command center and law enforcement agencies.

[0045] Specifically, by combining historical violation data and professional data, it is possible to formulate analysis and judgment rules more accurately, and use rich data sources to conduct trend analysis and event judgment, improve the comprehensiveness and accuracy of the analysis, extract traffic rules, laws and regulations and other information from professional data, formulate accurate analysis and judgment rules and operating parameters, ensure the legality and rationality of the analysis and judgment process, and use edge computing technology to achieve real-time processing and judgment of traffic analysis and judgment data, reduce data transmission and processing delays, and improve response speed. Trend analysis of historical violation data and comparison of legal basis and historical cases provide important background and basis for analysis and judgment, ensuring the reliability and accuracy of analysis and judgment. Real-time analysis and judgment provide dynamic law enforcement action suggestions, enabling traffic management and law enforcement departments to respond to traffic abnormalities quickly and flexibly and improve law enforcement efficiency. The platform reduces the burden of manual intervention and analysis through automated analysis and intelligent analysis, and improves the efficiency and accuracy of decision support. The transmission and sharing of real-time traffic analysis and judgment results enable the traffic command center and law enforcement departments to work together to optimize resource allocation and action strategies. Through accurate and timely identification of traffic anomalies and judgment of violations, the solution effectively improves the level of traffic safety management and reduces the possibility of accidents.

[0046] To sum up: by introducing the "end-to-end" concept in the field of computer science and information technology, and through artificial intelligence deep learning algorithms and big data mining technologies, the original data obtained by the front-end perception platform can be directly analyzed and judged without other preprocessing, reducing the requirements for data sources, data quality and data formats, etc., while supporting the scenario-based edge modular deployment and intelligent perception services of traditional perception facilities and cross-system and cross-platform result export and data resource sharing based on standardized interfaces. Finally, by switching the service areas and parameter adjustments of the application model, the output of law enforcement clues data and comprehensive analysis and judgment display in various comprehensive law enforcement fields can be established.

[0047] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0048] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. A digital traffic law enforcement insight engine and analysis platform based on edge computing, characterized by: include: Traffic data collection unit for: Deploy edge computing devices and sensing devices at traffic monitoring points. After deployment, different sensing devices will collect data. After data collection, traffic data to be analyzed will be obtained. Data fusion analysis unit, used for: The traffic data to be analyzed is fused. After the fusion is completed, edge computing technology is used to perform real-time data calculation. Edge computing technology is calculated through edge computing devices. After the real-time data calculation is completed, the traffic data to be predicted is obtained; Analytical data prediction and early warning unit, used for: The traffic data to be predicted will be used to predict abnormal events, and the degree of abnormality will be analyzed based on the predicted situation of the abnormal events. According to the analyzed degree of abnormality, an early warning will be issued at the traffic law enforcement terminal. Finally, the predicted abnormal events and the degree of abnormality of the abnormal events will be marked as traffic abnormality data.

2. The edge computing-based digital traffic law enforcement insight engine and analysis platform according to claim 1 is characterized in that: The traffic data collection unit is also used for: First, confirm the edge computing equipment and perception equipment. The edge computing equipment includes circuit boards, main control devices, AI chips, storage devices, interface devices, and temperature control devices; the perception equipment includes signal equipment, camera equipment, and radar equipment; Install the circuit board, main control device, AI chip, storage device, interface device and temperature control device in the designed chassis, fix the chassis in the predetermined position, and connect the power supply and network; The signal device is installed on the signal light pole at the traffic monitoring point, the camera device is installed at a position that can cover the entire traffic monitoring point, and the radar device is installed at a place that can detect the speed and position of the vehicle. The radar device and the camera device are used in conjunction with each other; When the sensing device collects data from the traffic monitoring points, each sensing device is first tested, and data collection is performed after all the devices are qualified. The traffic monitoring points include highway overload detection stations, dynamic monitoring points, expressway toll stations, freight sources, provincial boundary non-stop gantries, ferry crossings, comprehensive law enforcement workstations, etc. The signal equipment collects data on signal phase, signal timing and traffic violations; the camera equipment collects data on vehicle images, vehicle video streams, vehicle trajectories, traffic flow and average speed; the radar equipment collects data on vehicle speed, vehicle spacing and vehicle position; After all the data are collected, the traffic data to be analyzed are obtained.

3. The edge computing-based digital traffic law enforcement insight engine and analysis platform according to claim 2 is characterized in that: The radar equipment is used in conjunction with the camera equipment, including: The radar device collects vehicle speed in real time; When the vehicle speed exceeds a preset first vehicle speed threshold, the camera is triggered to collect vehicle tracking video streams for the vehicle; When the vehicle speed does not exceed a preset first vehicle speed threshold, but exceeds a preset second vehicle speed threshold, a vehicle corresponding to the vehicle whose speed exceeds the preset second vehicle speed threshold is taken as a target vehicle, and the target vehicle is marked; Retrieving the vehicle driving parameters of the target vehicle obtained by the radar equipment at each traffic monitoring point previously passed by the target vehicle, wherein the vehicle driving parameters include the vehicle speed of the target vehicle, the vehicle distance between the target vehicle and the rear vehicle, the vehicle distance between the target vehicle and the front vehicle, and the vehicle speeds of the rear vehicle and the front vehicle; Using the vehicle driving parameters of the target vehicle at each traffic monitoring point, obtaining a first vehicle driving safety factor of the target vehicle at each traffic monitoring point; The first vehicle driving safety factor of the target vehicle at each traffic monitoring point is obtained by the following formula: ； Among them, S 01 represents the first vehicle driving safety factor of the target vehicle at each traffic monitoring point; D x and D h V represents the vehicle distance between the target vehicle and the preceding vehicle and the vehicle distance between the target vehicle and the following vehicle respectively; x and V h are the vehicle speeds of the front vehicle and the rear vehicle respectively; V d Indicates the vehicle speed of the target vehicle; D a Indicates the preset distance reference value; V c represents a preset vehicle speed reference value; A represents a reference constant, and the value range of the reference constant is 1.5-100; Comparing a first vehicle driving safety factor of the target vehicle at each traffic monitoring point with a preset safety factor threshold; When the first vehicle driving safety factor of the target vehicle at any traffic monitoring point is lower than a preset safety factor threshold, the camera is triggered to collect vehicle tracking video streams for the target vehicle; When the first vehicle driving safety factor of the target vehicle at each traffic monitoring point is not lower than the preset safety factor threshold, the safety of the target vehicle is continuously judged, and when the safety of the target vehicle is lower than the safety requirement, the camera is triggered to perform vehicle tracking video stream acquisition on the target vehicle.

4. The edge computing-based digital traffic law enforcement insight engine and analysis platform according to claim 3 is characterized in that: When the first vehicle driving safety factor of the target vehicle at each traffic monitoring point is not lower than the preset safety factor threshold, the safety of the target vehicle is continuously determined, and when the safety of the target vehicle is lower than the safety requirement, the camera is triggered to perform vehicle tracking video stream acquisition on the target vehicle, including: When the first vehicle driving safety factor of the target vehicle at each traffic monitoring point is not lower than a preset safety factor threshold, extracting the first vehicle driving safety factor of the target vehicle at each traffic monitoring point; Obtaining a comprehensive vehicle driving safety factor using the first vehicle driving safety factor of the target vehicle at each traffic monitoring point; Extracting vehicle travel parameters of each subsequent traffic monitoring point after the vehicle speed of the target vehicle exceeds a preset second vehicle speed threshold, wherein the vehicle travel parameters include the vehicle speed of the target vehicle, the vehicle distance between the target vehicle and a rear vehicle, the vehicle distance between the target vehicle and a front vehicle, and the vehicle speeds of the rear vehicle and the front vehicle; Obtaining a second vehicle driving safety factor corresponding to the target vehicle at each subsequent traffic monitoring point by using the vehicle driving parameter of each subsequent traffic monitoring point after the vehicle speed of the target vehicle exceeds a preset second vehicle speed threshold; The second vehicle driving safety factor corresponding to each subsequent traffic monitoring point of the target vehicle is obtained by the following formula: ； Among them, S 02 represents the second vehicle driving safety factor corresponding to each subsequent traffic monitoring point of the target vehicle; D mx and D mh V respectively represent the vehicle distance between the target vehicle and the preceding vehicle and the vehicle distance between the target vehicle and the following vehicle corresponding to each subsequent traffic monitoring point; V mx and V mh Respectively represent the vehicle speed of the preceding vehicle and the vehicle speed of the following vehicle corresponding to the target vehicle corresponding to each subsequent traffic monitoring point; V d Indicates the vehicle speed of the target vehicle; D a represents a preset distance reference value; A represents a reference constant, and the value range of the reference constant is 1.5-100; using the comprehensive vehicle driving safety factor to compare with a second vehicle driving safety factor corresponding to the target vehicle at each subsequent traffic monitoring point; When the second vehicle driving safety factor corresponding to the target vehicle at each subsequent traffic monitoring point is lower than the comprehensive vehicle driving safety factor, the camera is triggered to perform vehicle tracking video stream acquisition on the target vehicle.

5. The edge computing-based digital traffic law enforcement insight engine and analysis platform according to claim 4 is characterized in that: Obtaining a comprehensive vehicle driving safety factor using the first vehicle driving safety factor of the target vehicle at each traffic monitoring point includes: Extract the vehicle flow and vehicle speed corresponding to each unit time of each traffic monitoring point; wherein the value range of the unit time is 10s-30s; Obtain a weight value corresponding to each traffic monitoring point using the vehicle flow and vehicle speed corresponding to each unit time corresponding to each traffic monitoring point; The weight value corresponding to each traffic monitoring point is obtained by the following formula: ； Among them, W j represents the weight value corresponding to the jth traffic monitoring point; n represents the number of vehicles running per unit time at each traffic monitoring point; C i V represents the traffic volume per i-th unit time corresponding to each traffic monitoring point; i represents the average vehicle speed per unit time corresponding to each traffic monitoring point; T i represents the traffic light control response delay rate of the i-th unit time corresponding to each traffic monitoring point; Obtaining a comprehensive vehicle driving safety factor by using the weight value corresponding to each traffic monitoring point and the first vehicle driving safety factor corresponding to each traffic monitoring point; The comprehensive vehicle driving safety factor is obtained by the following formula: ； Among them, S c represents the comprehensive vehicle driving safety factor; m represents the number of traffic monitoring points; W j represents the weight value corresponding to the jth traffic monitoring point; C pj represents the average traffic flow corresponding to the jth traffic monitoring point; S 01j Represents the first vehicle driving safety factor corresponding to the j-th traffic monitoring point.

6. The edge computing-based digital traffic law enforcement insight engine and analysis platform according to claim 2 is characterized in that: The data fusion analysis unit is further used for: The collected traffic data to be analyzed is preprocessed. The data preprocessing process includes data cleaning, data alignment, data conversion and data standardization; After data preprocessing is completed, time synchronization and space synchronization are performed. Time synchronization includes time standard unification, time alignment, missing time processing and time window division. Space synchronization includes coordinate system one, space alignment, missing space processing and space division. After the time synchronization and space synchronization are processed respectively, the time-space index is established and verified, and finally the traffic data to be analyzed after the time synchronization and space synchronization are completed are obtained, and the features of the traffic data to be analyzed after the time synchronization and space synchronization are completed are extracted; After feature extraction is completed, the feature data collected by different devices are associated and matched; The sensor fusion algorithm is used to fuse the associated and matched data.

7. The edge computing-based digital traffic law enforcement insight engine and analysis platform according to claim 6 is characterized in that: The data fusion analysis unit is further used for: The traffic data to be analyzed after data fusion is input into the edge computing device, and the edge computing device uses streaming data processing technology to process the input real-time data stream; The process of processing the input real-time data stream includes: According to the deep learning model deployed on the edge computing device, the input real-time data stream is inferred in real time, and AI inference data is obtained after real-time inference; Use GPU accelerated computing of AI chips to perform parallel and distributed computing on AI inference data; After parallel and distributed computing is completed, the traffic data to be predicted is obtained.

8. The edge computing-based digital traffic law enforcement insight engine and analysis platform according to claim 7 is characterized in that: The analysis data prediction and early warning unit is also used for: Obtain a random forest model and historical traffic data from a database, use the historical traffic data to train the random forest model, and obtain a standard random forest model after the training is completed; Extracting features for abnormal event prediction from the data to be predicted, the features including traffic flow change rate, vehicle speed abnormal value and signal phase duration; The extracted features are input into the standard random forest model to infer abnormal events. Before inferring abnormal events, the output inferred abnormal events are confirmed by setting a threshold. According to the abnormal threshold range of the inferred abnormal event, the abnormal degree of the abnormal event is confirmed, and the abnormal degree is divided into level one abnormality, level two abnormality and level three abnormality; Then, the warning level at the traffic law enforcement terminal is confirmed according to the degree of abnormality, and the warning levels are divided into level one, level two and level three. Finally, the inferred abnormal events and the abnormal degree of the abnormal events are uniformly labeled as traffic anomaly data.

9. The edge computing-based digital traffic law enforcement insight engine and analysis platform according to claim 1 is characterized in that: Also includes: Data integration unit for: Integrate traffic anomaly data with analysis information, and obtain standard traffic analysis data after data integration; Among them, the abnormal data in the traffic abnormal data is extracted, and the abnormal data includes the abnormal event type, abnormal degree, occurrence time and event location; Confirm the assessment information, which includes weather information and special time information; The traffic anomaly data is integrated with the judgment information of the same time and place. At the same time, the characteristics of traffic anomalies are combined with the characteristics of weather and special time information to form a comprehensive feature set; According to traffic management and law enforcement rules, establish the association rules between traffic anomalies and analysis information; Associating association rules with comprehensive feature sets, and verifying accuracy and completeness after associating information; After accuracy and completeness verification is completed, standard traffic analysis data is obtained.

10. The edge computing-based digital traffic law enforcement insight engine and analysis platform according to claim 9 is characterized in that: Also includes: Integrated data analysis and decision-making unit, used for: The standard traffic analysis data is analyzed and judged using historical violation data and professional data, and finally the analysis and judgment results are transmitted to different processing centers; Extract historical violation data and professional data from the database. Historical violation data includes violations, penalty records and accident situations; professional data includes traffic rules, laws and regulations, traffic flow statistics and historical accident analysis; Formulate research and judgment rules based on professional data, including the association rules between abnormal events and illegal behaviors, and convert the formulated research and judgment rules into operational parameters; Conduct similar trend analysis on key feature data in standard traffic analysis data based on historical violation data. Key feature data include abnormal event type, abnormal degree, occurrence time and location; Use professional data to provide regulatory basis and historical case comparison for standard traffic analysis data completed for similar trend analysis; After the legal basis is provided and compared with historical cases, the traffic analysis data to be judged is obtained; Input the traffic analysis data to be judged into the analysis model for real-time analysis and judgment; Real-time analysis and judgment is to judge the traffic analysis data according to the analysis rules and make law enforcement action judgments; After real-time analysis and judgment of the traffic analysis data to be judged, the real-time traffic analysis result is obtained; Finally, the real-time traffic analysis results are transmitted to the traffic command center and law enforcement agencies.

Images