STUDY ON PREDICTION METHODS FOR SUBGRADE SETTLEMENT OF HIGH-SPEED RAILWAY USING GRAY NEURAL NETWORK

Prediction methods for subgrade settlement are an important issue in the construction of high-speed railway, grey theory GM(1, 1) and BP neural network model are widely used to predict. These two models have unique advantages compared to common models, despite of limitations. So we combined with the advantages of the grey theory GM(1, 1) model and BP neural network model in this article and put forward a new linear combination forecast model. The data measured in Tianjin-Qinhuangdao Passenger Dedicated Line subgrade settlement is analyzed with this new model in this paper. The results show that both the advantages of the GM(1, 1) model and BP neural network model are absorbed in this new combination model and it has higher prediction accuracy. Thus, its wide application prospects can be expected.