Application of fuzzy neural networks for predicting seismic subsidence coefficient of loess subgrade

Taking Zhengzhou-Xi'an passenger dedicated line as an example, based on the analysis of the main influencing factors, a fuzzy neural networks model for predicting seismic subsidence coefficient of loess subgrade has been established. The model combines the fuzzy information optimization technology and neural network. It integrates the two theories, by making up the defects of the neural network in fuzzy data processing and the deficiencies of fuzzy logic in learning. The results show that model is quite suitable to predict the seismic subsidence coefficient.