"OD Matrix Estimation" Technology In Urban Traffic Demand Forecast Prediction Research

Traffic demand forecast is an important part of traffic planning and crucial technical content, and "Four-step" traffic planning method, the theory of which has been widely used in China, is the classic method of traffic demand forecast. But in recent years, the work practice of urban Traffic planning in china shows that classic Four-step planning is limited in our country nowadays. To this problem, the thesis suggests the"OD matrix estimation"technology in urban Traffic demand forecast prediction research, and have research, the main research contents are as follows:First, the thesis analyzes the main Features of traditional four-step method and at present the main problems, and the traditional "four-step" at all stages of predictive models in the practical application of reliability have been analyzed, the traditional Four-step planning is limited in theory and practice from the analysis of the traditional "four-steps "prediction model limitations in our current application, the thesis Have brought forward necessity of"OD matrix estimation"technology in urban Traffic demand forecast prediction research.The thesis studies the Summary and evaluation, of OD matrix estimation on the base of domestic and foreign OD matrix theory, and have been studied in obtaining basic data of OD matrix estimation, as well as Analysis of the "OD matrix estimation" factors affecting the accuracy of test methods. The thesis suggests that the OD matrix estimation method based on TransCAD, and has studied for "OD matrix estimation "process and data preparation, especially for obtaining the initial OD matrix, network traffic observation and other crucial issues.Finally, the context with relevant examples analyses the feasibility and effectiveness of"OD matrix"and the steps and methods of the traffic demand prediction. And, at the same time, it also determines the key steps of the fundamental data preparation, the OD matrix, and the calculation of traffic generation.