Research On Coal Railway Logistics Demand Forecasting Models

China being as a developing country, the coal will continue its guarantee for high-speed economic growth in our country for a long period. Coal logistics is an important part in the coal industry. Transportation is the core of coal logistics, transport and distribution. In our country, more than60%of the coal is transported by rail, but it seems to be inadequate for railway transportation. By means of the qualitative and quantitative analysis method, the model is established to forecast accurate demand quantity with the well grasp of the main influential factors in coal railway freight volume, hoping to provide some important bases for the government to perfect relevant plan, infrastructures and logistics supply system.There are mainly three parts in this thesis. In the first part, after analyzing the importance of demand forecasting in coal railway logistics, coal railway freight volume are chosen as the representation of coal railway demand. The combination qualitative analysis and quantitative calculation of grey correlation degree is used to choose the affecting factors of coal railway logistics demand. In the second part, four kinds of coal railway logistics demand forecasting modeling method being as PSO-SVR forecasting model, Grid search-SVR forecasting model, BP neural network forecasting model and combination forecast model of railway logistics demand are put forward. In the last part, with the help of the above models, railway freight volume during1995is forecast and the results of the various prediction methods were analyzed. In the end, Combined with the12th five-year plan by the national bureau of statistics, the ministry of railways and energy administration departments, extrapolation prediction is calculated. And the result is put into the highest precision of PSO-SVR model to forecast the coal railway logistics demand in2012-2015. Then suggestions are put forward to the development of railway logistics of coal.