Identification And Prediction Of Urban Traffic Congestion

At present, the traffic congestion exist in large and medium-sized city all over theworld. Urban traffic congestion has serious dangers, its direct consequence is mainlymanifested in two aspects: First, the urban traffic congestion caused significant time delay andenergy waste, and brought great economic losses to the society. Second, the urban trafficcongestion make very low speed of vehicles. Low speed vehicles will emit large quantities ofwaste and produce a lot of noise, so that the quality of the urban environment greatlydecreased, thus causing serious harm to the health of urban residents. If can be timely andaccurately identification and prediction has been formed or to be formed the urban trafficcongestion, develop reasonable and effective strategy to ease traffic congestion. It will begreatly reduce the negative effects caused by traffic congestion and greatly improve theoperation level of the whole city road traffic system. Therefore, study of identification andprediction of urban traffic congestion is necessary and meaningful.This paper propose a new urban traffic congestion recognition algorithm. In-depth studyshort-term forecasting of traffic flow parameters, which is the key technology of urban trafficcongestion prediction. Establishment of a multi-model fusion forecasting methods. The maincontents of this paper include the following aspects:First of all, study the operating characteristics of the traffic flow. Study the definition,classification and metrics of traffic urban congestion. And analyzes the deep level reasons ofthe formation of urban traffic congestion.Then, analyze and summarize the classic method of traffic congestion identification basis,and put forward a kind of identification algorithm of urban traffic congestion based onupstream inflows flow rate and downstream outflows flow rate. Using the real traffic datacome from Xiaozhai corss in Xi’an, verification and analysis about this algorithm.Finally, In-depth study short-term forecasting of traffic flow parameters, fusion ofseveral commonly used short-term traffic flow prediction method, establishment of amulti-model fusion forecasting methods. Using the real traffic flow data come from southsecond ring road in Xi’an, verify and analyze about two commonly used methods and multi-model fusion forecasting method.