Research On Knowledge-Based Methodologies And System For Decision Making Of Traffic Congesion Management

Aiming at constructing an intelligent decision support system with data driven decision support methods for urban traffic congestion management, several key issues are discussed in this dissertation.At first, in order to make decision more effectively and sufficiently, the elemental characters and main causes of urban traffic congestion are analyzed. Especially, for its spatio-temporal peculiarity, a real time analytic method to model urban congestion's spatio-temporal peculiarity and its trend is presented. And then a universal model describing the urban traffic congestion is brought forward. At same time, some essential relations between congestion types and its dispersing strategies are drawn out and a basic description of the decision problem, the decision process and some special characters of this process are presented. On the Basis of those analyzed before, the necessity of applying knowledge-based system in the urban traffic congestion dispersion decision process and what kinds of knowledge are necessary during the decision process are discussed.Secondly, based on the analysis of the data involved in the urban traffic congestion management decision process, the data flow during making decisions to disperse urban traffic congestion and how to manage those data efficiently are studied. For the purpose of further decision and data mining, a data warehouse based data model is presented.Later on, by importing rough set theory and method into the urban traffic congestion management decision process, the relations between the real time data drawn from the urban traffic monitoring system and the severity levels of the urban traffic congestion is studied, a knowledge-based model to deal with the uncertainty in urban traffic congestion pattern recognition is constructed and then a algorithm based on rough set theory to compute the weight of each aspect of a case which will be used while dealing with the traffic alarms through case-based reasoning method is presented.And then, via extending the traditional quantitative decision model to knowledge based decision model, cooperating qualitative reasoning methods with quantitative computing method, integrating data warehouse technique, case based reasoning method and knowledge acquisition method into the congestion dispersing process, the idea of applying knowledge based decision support system to deal with urban traffic congestion is brought around, including the decision support system's structure and its analytical model.At last, the stepwise realization of the knowledge-based urban traffic congestion dispersing decision support system armed with rule based reasoning ,case based reasoning and knowledge acquisition is showed.