The Research On The Combination Problem Of Railway Volume In Our Country

With the development of Chinese economy, the improvement of people’s living standards as well as the continuous improvement of the railway construction, people’s consumption concept has changed. Traveling tours, visiting relatives and other activities is becoming a very important part of people’s lives, therefore the railway passenger volume also increased with the increase of people’s long-distance traveling. How to forecast railway passenger volume scientifically in order to assist the railway sector to determine the macroeconomic development strategy of railway, control and allocate passenger volume to ensure the smooth flow of railway operation, has a very important significance.In this paper, we’ll have a comparatively detailed and deep study about the forecasting of Chinese railway passenger volume. This paper first analyzes the development of China’s railway transportation system, presents the importance of accurate prediction of railway passenger volume. Followed by analyzing the study status at home and abroad and summarize the main ways to predict railway passenger volume accurately. And then introduced three single forecasting methods used in railway passenger volume prediction:multiple regression analysis prediction, time series prediction, support vector machine regression analysis. For multivariate regression analysis, this paper introduces the principle and modeling steps of multivariate regression; for time series analysis, this paper first introduced the concept and nature of stationary time series, its statistical properties and modeling steps, finally analyzed the implementation process of summing the autoregressive moving average model; while for support vector regression machine, we first introduce the case of support vector classification machine linear separable or not, then put forward the regression function of support vector machine. Then this paper puts forward three kinds of methods variance given different weights of the reciprocal method. Finally, using three methods of passenger traffic volume prediction of railway in our country to conduct the empirical analysis, and compare the results of these three methods with combination forceasting results,and lastly get the conclusion that the accuracy of combination prediction is higher than the single method.