Traffic Anomaly Analysis Based On Electronic Police Data

With the rapid growth of car ownership,the traffic safety situation in urban roads is becoming more and more serious.People are also paying more and more attention to traffic safety issues.One of the fine and quick traffic jam judgment and traffic safety are closely related.On the one hand,congestion exceeding a certain level can easily lead to accidents.On the other hand,unexpected traffic congestion caused by traffic accidents can cause secondary accidents if it is not discovered promptly.The traditional traffic condition detection relies on various types of vehicle detectors such as coils and radars to obtain the traffic and the point speed of the road section,can not obtain the source and destination of the vehicle,and can not obtain the information about the travel time of the vehicle and the delay of the traffic light.With popularization and application of electronic police,intelligent monitoring and recording system,GPS,electronic license plate and other technologies,the vehicle identity in the city can be accurately identified and the running track of the vehicle in the city can be obtained.Spatiotemporal multidimensional features of trajectory data provide more data support for finely determining traffic states,especially accidental ones.Based on the license plate data obtained by electronic police equipment at 107 intersections in Suzhou Industrial Park,this paper proposes to increase the license plate matching rate by using license plate fuzzy matching based on the edit distance algorithm to improve the utilization of license plate data.The vehicle matching data is used to judge the traffic status of road sections,intersections and the whole road network,and the existence of the macro basic map in the real road network is analyzed.Summarize the general law of the state of road network in one day,and finally judge whether there is sporadic congestion in current road network.