Research On Queue Length Of Traffic Flows At The Intersection Based On Electronic Police Data

The detection of the queue length at road intersections can effectively monitor and prevent traffic congestion,and is a major theoretical basis for alleviating the pressure on the road network and promoting the healthy development of urban traffic.In the past,GPS data and cross-section detection data had defects such as low sampling rate and insufficient information content.The judgment of the queue length was biased and the representation of the queue mode was not precise enough.In recent years,the bayonet-type electronic policy has developed rapidly,and the coverage rate has increased year by year.Through the mutual comparison of license plate recognition,refined vehicle trajectories and inter-vehicle relationships can be obtained,providing a solid basis for the precise control of traffic signal control equipment.Critical data support in urban transportation research.This article mainly focuses on the electronic police data to study the traffic queue length at intersections,which is divided into three parts:(1)Refined and trajectory processing of electronic police data,and analysis of the distribution characteristics of vehicle travel time based on the full sample of electronic police data;(2)According to the number of times the vehicle has stopped on the road segment,a formula for calculating the maximum queue length experienced by the vehicle is proposed;(3)prediction of vehicle travel time and queue length based on deep learning algorithms.Combining the example data of Ningbo city to verify and evaluate the different performance of the queue length model and the prediction model in the morning and evening peak and flat peak periods.