Research On Passenger Volume Predict Of Passenger Dedicated Line

The Donghaidao Shinkansen in Japan, which was the first high-speed railway in the world, was put into operation in1964. With the advantages of huge transport capacity, high speed, safety and punctuality rate, high-speed railway has been increasingly developed worldwide. In order to develop the railway, enhance the competitiveness in passenger transport market, relieve the pressure of existing railway, China railway has constituted the medium and long period layout of the railway network, which elaborates that our country has access to the period of rapid development of Dedicated Passenger Line (PDL).As the large-scale construction and successive operating of PDL in transport corridor, the transport pattern and structure of transport corridor have been changed. Meanwhile, the passenger volume of PDL determines not only the investment of construction projects, but also passenger train plans, transportation organizational modes and marketing strategies of passenger transport. Therefore, on the basis of available data of transport corridor, it is provided with great significance that how to realize a scientific and reasonable prediction method of the passenger volume of PDLFirstly, in this thesis, the significance, main content and technology roadmap of predicting the passenger volume of DPL are described. It particularly introduces and analyses the prediction methods for railway passenger volume at home and abroad, in accordance with the advantages and disadvantages of which, proposing a new prediction method for the passenger volume of DPL based on the improved four stager manner Then, the influencing factors of the passenger volume of transport corridor with the purpose of selecting the main influencing factors are qualitatively analysed and quantitatively calculated. What’s more, in light of the formation mechanism of the passenger volume of transport corridor, which is classified as trend volume and induced volume, the thesis designs and constructs BP neural network model and induced coefficient model to predict the whole passenger volume. Subsequently, the generalized cost and normal distribution of time value are used to reckon up the sharing rate of passenger transport to predict the passenger volume of DPL on the basis of time-value model in the thesis. At last, the thesis takes Liuzhou-Nanning Dedicated Passenger Railway as an example to test and verify the rationality of the prediction method proposed.