Research On The Key Technology Of Urban Road Traffic Flow Prediction And State Identification

With the sustained steady and rapid development of China’s economy, the pace ofurbanization is accelerating, the contradiction between the rapid development of citytraffic demand and urban road traffic infrastructure supply at a slow pace is increasinglyprominent, the problem of traffic congestion has become a serious obstacle to the sustainabledevelopment of city, and bring serious influence to people’s daily life and work. Thedevelopment of domestic and international practice shows that, through more informationaland intelligent management system to improve the level of traffic management based onexisting transport facilities, is an effective means to solve the problem of urban road trafficradically. The goal of Nansha district intelligent traffic management and controlplatform project is to raise the level of management, to guarantee the the smooth of urbantraffic flow, to improve traffic order and to enhance the level of service. Accurate predictionof urban road traffic flow and identification of traffic state are important parts of thissystem. The research content of this paper is from Nansha district intelligent trafficmanagement and control platform project.This paper is focused on the prediction of short-term traffic flow and identification ofroad traffic state. The purpose of this study is to provide strong technical support forintelligent traffic management and control platform, make the management more scientific.The research mainly includes the following aspects：According to the need of the prediction of traffic flow and traffic statediscriminant, traffic data preprocessing method is proposed. The article briefly introduces therecognition and processing technologies of fault data. Through the pretreatment ensured thequality of traffic data, which guarantee the accuracy of the result of traffic flow predictionand state identification.A large number of studies have shown that it is difficult to satisfy the demands ofexpected accuracy with a single forecasting model. The article first introduced the principle ofBP neural network for short-term traffic flow prediction. Describes the modeling process andgives the improved method aiming at its shortcomings. Then establish short-term traffic flowforecasting model based on WNN. Multi-model fusion prediction algorithm based on datafusion is put forward finally. The effectiveness of multi-model fusion prediction algorithm isverified through MATLAB simulation analysis.. In this paper, we determine the road traffic state by means of fuzzy comprehensiveevaluation because of the fuzzy uncertainty of road traffic state. In order to reduce the influence of subjective factors, we determine the weight coefficients with the fuzzy AHP.Finally the feasibility of the algorithm is verified by the example analysis.On the basis of Nansha district intelligent traffic management and control platform, thepaper introduced the implementation process of traffic flow forecasting and traffic stateidentifying.It provides reference for large and medium-sized cities to solve related problems.