| Combined with the sub-project "Basic Traffic Information Collection and Fusion Technical Research" of National 15 science and technology significant project "Intelligent Traffic System Crucial Technical Development and Demonstration Project" etc., this dissertation is completed. The dissertation consists of six chapters.The first chapter introduces the research background, purpose and content of this dissertation. With the appearance of Intelligent Traffic Systems and increasing deterioration of traffic congestion problem of all the countries, traffic managers have realized that Advanced Traffic Management Systems is the most effective measure of solving traffic congestion presently. As one of ATMS's crucial technologies, the research of traffic congestion automatic identification algorithm has already begun in 1960's, such as famous California algorithm and McMaster algorithm. Some algorithms have applied into practice, and It's proved that they are being not identical in identification performance because of mostly the limitation of detection technology and data analysis technology of that time. This is also one of major background of author researching new algorithm. That effective enforcement of traffic congestion automatic identification algorithm makes the traffic management department can real time and effectively trail traffic state of highway, once discovering the traffic state occurring abnormity, they should adopt prevention and navigation measures promptly. Through information dissemination system,the congestion condition will be provided to travelers, and avoiding second congestion to make road network serious seriously obstructed. Therefore effective traffic congestion identification algorithm has significant economic benefit and social significance for solving the traffic congestion problem on highway.The second chapter has analyzed the property of traffic congestion in detail, and has qualitatively described the change regulation of three basic traffic flows parameters under two type traffic congestion patterns. Based on this, this dissertation has simulated different level traffic congestion using VISSIM traffic simulation software, and has analyzed the applied condition of ACI algorithms. The dissertation has analyzed the relation between SDP,FDP and traffic demand, capacity, detection station distance under different congestion condition using simulated data, and has elicited following conclusion:1. The validity of traffic congestion data is concerned with the distance of detection station which neighbors on congestion, and exists a best range distance.2. The change level of traffic parameters is concerned with road traffic capacity dropping level. When the capacity drops to certain level, the change of traffic parameters would be obvious comparatively.3. Under condition of small traffic flows, the little fluctuation of traffic flows can arouse the obvious fluctuation of detection station data.At last, this chapter has given system frame of highways traffic congestion state identification.The third chapter has analyzed the needed parameters of ACI algorithms and the acquirability of parameters. Firstly, the dissertation has analyzed the principles of choosing traffic parameters of ACI algorithm, and from two aspects of traffic management department and classical ACI algorithm, dissertation has analyzed general traffic parameters. Then, based on pre-construction demand of highways ACI algorithm design, this dissertation has analyzed the needed traffic information, especially dynamic traffic information. Combined with existing traffic collection technology, dissertation has introduced the traits of different detectors and its supplied parameters, which has also offered parameters option for the input variable design of algorithm.The fourth chapter has analyzed the design process of highways ACI algorithms. Firstly, this dissertation has looked back classical ACI algorithm, and has analyzed these algorithms' shortcomings on identification performance, and has discussed the... |
