Research & Development On Theory And Methods Of Traffic Volume Forecast System

Energy sources and transportation, both of which are closely relative with transportation planning, will be two factors influencing the economy development in the near future. Traffic investigation and demand forecast are the very foundation for the transportation planning and an efficient and practical forecast system dealing with large quantity of information is necessary for the prediction of the traffic demand and its development as well as the decision of the transportation infrastructure layout, which is urgent to the reconnaissance and design academy, scientific research institute and some corporations. Chinese and foreign experts and scholars did a lot of researches on traffic demand forecast and traffic planning. However, these researches are not so practical. Based on the existing researches, the thesis discusses the feature, influencing factors, and thought of the traffic demand, analyzes the basic principles and main methodologies, constructs and develops the TVFS V1.0 forecast system and applies it to the scientific researches, which have proved it to be effective and reliable. The thesis consists five chapters in five parts.Firstly, the author analyzes the traffic demand, traffic generation, distribution, modal split and assignment, programs four-step forecast models, defines the data structure and model algorithm, provides 16 models for different forecast phases, and analyzes the parameters, solution and test. Based on software engineering and windows platform, the author develops TVFS V1 .0 forecast system, which can be applied to project management, data processing, forecast preparation, forecast model, graph display, integrated cost calculation, online help and output.Secondly, the author describes professional software engineering management in developing the TVFS V1 .0 forecast system so as to ensure the process and quality of software development and improve the efficiency, analyzes the demand of the software as a professional traffic demand forecast system, provides object oriented modeling and designs the software system framework and DDL for the forecast functions to enhance the flexibility. The author also defines the data structure used in the models, which realizes data sharing with other forecast systems and databasesand broaden the use scope of the software.Finally, the author analyzes the objective of transportation planning, and introduces an example of an real scientific research project, which compares the results of using TVFS and without it so as to validate the feasibility and efficiency of the theory and methodology.This thesis proposes and realizes the combination model of modal split and traffic assignment base on multi-modal and multi-path theory, introduces dummy nodes, integrated link impedance and transfer impedance to solve the modal split problem and applies it to real planning project. The exploration and development enriches the theory and models of traffic demand forecast and is of great importance to the traffic planning improvement and traffic planning, inter-city rail transportation and road network planning carried out in China.