Evaluation And Analysis Study On Operation Status Of Freeway Section

In recent years, along with the rapid development of freeway in our country, the freeway network is developing gradually. At the same time, freeway management has to be face with higher requirement and greater challenge than before. The purpose of this dissertation is, through the study on comprehensive evaluation and analysis of operation status of freeway section, to provide the scientific bases and datum for freeway governor, and to improve the level of freeway network management and service, further. Considering the actuality of freeway in our country, this dissertation is studied for three important aspects on freeway section: dynamic comprehensive evaluation of operation status, automatic incident detection and shot-term traffic flow prediction. The three aspects are related to one another, and each aspect reflects the dynamic operation status of freeway section from a special side.In the first aspect, this dissertation firstly presents the evaluation attribute system based on analyzing common attributes, then the dissertation study and analysis all existing comprehensive evaluation methods. According to the requirement of dynamic comprehensive evaluation of operation status on freeway section, the dissertation presents a new dynamic evaluation method called"classify status first, evaluate trend later". It is the first time that the dynamic comprehensive evaluation method is applied on system evaluation of traffic field. The application example proves the validity of the new method.In the second aspect, the history and development of incident detection algorithms are firstly reviewed briefly, and then the theories and characteristics of all popular incident detection algorithms are analyzed. Based on the analysis of their advantages and disadvantages, and considering the actuality of freeway in our country, a more practical and effective incident detection method is presented.In the third aspect, the dissertation presents a short-term traffic flow prediction model based on improved back propagation (BP) neural network after analyzing the existing traffic flow prediction methods. Adaptive study rate and Levenberg-Marquardt training algorithm are applied for the improvement and optimization of BP neural network. At the end of this chapter, simulation experiments of this algorithm are illustrated, the result of which proves the feasibility and efficiency of the algorithm and high application value in traffic flow prediction.This dissertation is supported by the scientific research project that Analysis Method and Application Study on Operation Status of Freeway Network in Jiangsu Province. The research achievements are practical and applicable.