Railway Freight Volume Forecasting By Artificial Neural Network Based On Economic Cycle

As one of the most important means of transport in the comprehensive transportation system, Railway transportation plays an irreplaceable role in the long-distance transportation of bulk-goods and makes great contributions for the high-speed-steady economic development. Railway freight volume forecasting is the basis for railway development plan and the premise for operating decisions of transport enterprise operating decisions. Scientific freight volume forecasting is very important for making decision of railway development strategies and giving full play to railway transportation capacity. But the railway transportation system is a complex dynamic system affected by many factors in which the most important influence factor is national economy macro-environment. Most of the traditional forecasting methods that ignore the influence of national economy macro-environment have the inadequacies of the unexpected errors and weak anti-jamming capability. So, it is necessary to study the influence of national economy macro-environment and build a more reasonable model for railway freight volume forecasting.First, the influence factors of railway freight volume and the advantages and disadvantages of different methods of railway freight volume forecasting at present were analyzed based on researching references at home and abroad.Then, the viewpoint that the unexpected errors can only be reduced and the anti-jamming capability of forecasting can only be improved if the influence of the economic cycle was fully considered was made based on analyzing the economic cycle theory and its influence on railway freight volume forecasting. Meanwhile, the length of time on training sample was determined as the recent complete economic cycle when using artificial neural network for railway freight volume forecasting.Based on the theory of economy cycle and artificial neural network, the concept of economic-cycle-phases parameter was put forward and the railway freight volume forecasting model by economic cycle was established. The model consists of the factors analysis module, economic cycle analysis module and integrated forecasting module. As the core of the model, the accurately measurement and division of economic cycle is the premise of precise forecasting.Finally, the improved forecasting modal was test by the data of Datong-Qinhuangdao railway freight volume and compared with the traditional model. The results show that the improved model is more effective and dynamic network is feasible in railway freight volume forecasting.