Research On Traffic Incident Detection For Beijing Expressway Based On Analysis Of Quantity And Quality

With the rapid development of economy, the whole society is increasing demand for transportation, the rapid growth of the traffic worsening traffic on the road, urban road traffic incident also occurred frequently. Researching traffic incident detection is so significant to improve the condition of urban expressway traffic, play to the superiority of the expressway and improve the management level of expressway.This paper proposed research on urban expressway traffic incident detection based on quantity and quality analysis by using the data collected from the second ring road of Beijing detector. First of all, the paper analyzes traffic flow temporal data; macro regularity can be drawn by analyzing the traffic flow. And then, the article discusses the analysis of traffic flow quantity identification and analyzes traffic flow data on one of the sections on the second ring road by using binomial probability model and output the hypothesis testing results by using Cars Mar application and MATLAB software; Secondly, the paper analyzes the traffic flow quality detection on the second ring road in Beijing, calculate the number and position of the mutation by using the least square model and MATLAB software, further analyze the section of road by using the iterated counting method which exhibit the change point, at the same time, the paper output the hypothesis testing results to determine traffic flow conditions. The paper analyzes the traffic flow quantity and quality detection, set up the traffic incident detection algorithm base on wavelet analysis theory and traffic flow characteristics. The paper decomposes the traffic flow data by using the MATLAB software and analyzes the flow parameter through wavelet decomposition coefficient. Finally, the paper output the results of data fusion by using MATLAB software and introduces the road service level to validate the results.Through the wavelet analysis, traffic incident detection based on quantity and quality has a better comprehensive detection performance in terms of detecting location and time of traffic flow mutation and traffic incident. At the same time, the method also provides the ideas and methods to establish a high performance of traffic incident detection algorithm.