Prediction Of Railway Passenger Volume Traffic Volume Basedd On Genetic Algorithms And BP Neural Network

In recent years, railway transportation of China has developed rapidly, numbers of railway engineering projects are under building or are going to be carried out in succession. Fortune capacity forecasting work is in the very important position in the earlier stage work of traffic engineering construction. Accurate consequence of passenger traffic volume prediction is one of major basises for traffic planning.Prediction accuracy is affected by Prediction Method. The aim of this paper is to apply GA-BP on Railway passenger traffic volume prediction which is formed by Genetic Algorithm and BP neural network. The goal of this paper is to find out the way to increase accuracy of prediction and study the prediction method of GA-BP algorithm.This paper firstly introduces the importance of Railway passenger traffic volume prediction and analyzes the methods on present and some new prediction theories, such as Prior moving average model, Exponent smooth model, GM (1,1) model which belong to time series, Regression Analysis model which belong to Influence Factors model and also analyzes the advantages, disadvantages and different scope of application.Secondly this paper studies China's railway passenger traffic volume trends and analyzes the impact of Railway passenger traffic volume-related.What's more, the article points out eight factors which affect the Railway passenger traffic volume according to the correlation coefficient and the selection criteria.Then, in this paper the Genetic Algorithm is used to optimize the connection weights of BP neural network to form GA-BP Algorithm as main prediction method. More over the article construct the model of causal relationship and time series.Followed that, by comparing the prediction result of causal relationship GA-BP model with BP model, the results show GA-BP overcomes the shortcomings that BP algorithm is usually trapped to a local optimum and is affected by the initializing weights and increases the accuracy of the prediction. By comparing the prediction result of causal relationship GA-BP model with common prediction method such as Exponent smooth Regression Analysis,GM(1,1), the results indicate that error precision of GA-BP algorithm is superior and GA-BP Algorithm is a available model prediction method. By comparing the prediction result of time series causal GA-BP model with relationship GA-BP model and BP model, the results show that time series causal GA-BP model disagree with the prediction of China's railway passenger traffic volume.In the finality, some points and the limitations of prediction of GA-BP model along with further studies are discussed.