The Prediction of Derailment Coefficient Based on Neural Networks

Derailment coefficient is an important criterion to evaluate the operating safety of rail vehicles. A derailment coefficient prediction method based on neural network is proposed in this paper. First, the basic concepts of derailment coefficient are briefly discussed. Then the principle of BP and NARX networks and their related learning rules are presented. BP network is compared to NARX network and their disadvantages are outlined. Finally, BP and NARX neural networks are established to analyze their prediction performances. The experimental result shows that, compared with BP neural network, NARX neural network offers better predictive performance of the derailment coefficient.