Study On The Critical Techniques Of Freeway Incident Management System

Combined with the sub-project "Basic Traffic Information Collection andFusion Technical Research" of National 15 science and technology significantproject "Intelligent Traffic System Crucial Technical Development andDemonstration Project" etc., this dissertation is completed. The dissertationconsists of seven chapters.The research background, purpose and content of this dissertation areintroduced firstly. With the development of society, freeway is becoming more andmore important. The administrator has realized that ATMS is the most efficient wayto solve traffic problems. The history and present situation of Freeway IncidentManagement System all over the world are reviewed, the experiences,achievements and the problems needed to solve are summarized.The composition, management step and meaning of FIMS are given, and thenthe demands of system are analyzed. With the demands, function which fit thedifferent parts is developed, so structure of system is constructed. Learning foreignadvanced technology, combining our nation's traffic management and developmenttrend, this paper proposes the new FIMS with Chinese feature.The third part introduces situation of AID in FIMS and ITS especially. Thetraditional AID algorithms have several limitations. They can not get satisfactoryDR and FAR at the same time, so the new ones should be developed. A newincident detection algorithm for freeway is developed based on the artificial neuralnetwork method by comparing and analyzing the incident and incident-freeconditions. The inputs of algorithm are designed according the traffic flow'scharacteristics in up and down streams. This proposed algorithm is tested by a largeset of simulated data. The results shows that the performance of this algorithm isbetter that some other algorithms already existed. Furthermore, when the algorithmcalibrated with the simulated data from one detection zone is used to another...