The Research On Models Of Traffic Generation And Distribution Prediction And Application Of Intercity Rail

The wave of economic globalization and regional integration has made China accelerated the pace of urbanization, the Yangtze River Delta, Pearl River Delta and other large urban agglomerations has formed. The continuous development of urban agglomerations makes that intercity passenger traffic has increased dramatically, then traditional traffic has been unable to meet the demand for intercity passenger travel, so the inter-city rail transport with such as fast, convenient, comfortable and "bus" features began play an important role. The prediction accuracy of traffic generation and distribution prediction has a direct impact on the rationality of inter-city rail transit planning and the social and economic benefits after it is operated. Domestic, the study of "four-stage method" is not yet mature, and some foreign prediction model does not apply to our transportation system, so this thesis is to study a more appropriate forecasting model to improve forecasting accuracy. In this dissertation, the main research results are as follows:Firstly, urban agglomerations and inter-city rail transport are described in detail, including the concept of urban agglomeration, formation conditions, the formation process, the definition, characteristics and intercity passenger features of inter-city rail transit. Combined with the relationship of intercity rail transportation and other major transportation way, the functional orientation of intercity rail transportation is given.Secondly, this thesis analyzes the influencing factors of intercity rail passenger systematically, and uses gray correlation analysis to determine the main factors. Based on introducing traditional traffic generated forecast models in detail and analyzing the scope and deficiencies of these models, the gray GM(1,N) model of intercity rail passenger forecasts is introduced. For the deficiencies of GM(1,N) model, the dissertation proposes the gray GM(1,N|Ï„,r) model to forecast the intercity rail passenger generation and analyzes the applicability of this model on generated prediction of intercity rail.Then, this thesis introduces the basic idea of distribution prediction and the traditional distribution prediction model in detail. By comparing the advantages and disadvantages of the traditional model, we determine the dual constraints gravity model is more suitable for intercity rail passenger distribution prediction. But due to the traditional dual constraints gravity model not considering the impact of industry related degree on inter-city passenger distribution, this dissertation proposes the improved dual constraint gravity model based on industry related degree, and presents the method for finding the model parameters in detail.Finally, the thesis makes the empirical analysis for an example of Wuhan city circle, and gets the passenger generation and distribution of five cities of Wuhan city circle. Comparing the GM(1,N|Ï„,r) model with multiple regression model and the GM(1,N) model, the dissertation gets the conclusion that the GM(1,N|Ï„,r) model has higher predicts accuracy on generated prediction of intercity rail. At last, the prediction distribution of passenger of part cities of Wuhan city circle is made preliminarily.