GA-BP Neural Network Prediction Model for Tunneling Speed of Shield Machine with Composite Formation Dual Mode (TBM-EPB)

Abstract

Relying on the construction site data of Shenzhen Metro Line 13 tunnel project, in this paper, the genetic algorithm (GA) was used to optimize the BP neural network (BP NN) to establish a new GA-BP NN prediction model to predict the tunneling speed of the shield machine. In order to obtain more reasonable and reliable tunneling parameters during the shield tunnel construction .Firstly, the influencing factors of the tunneling speed of the shield machine in the two modes were analyzed, and the sample data of relevant factors was established for this project; secondly, BP NN was trained with 740 sets data to build a prediction model, which was tested with 120 sets of data; then used genetic algorithm to optimize the BP NN, a mature trained GA-BP NN prediction model was obtained through network training; finally, the predicted and measured values of the two prediction models were compared and the errors are analyzed. The results show that the optimized BP NN model has higher accuracy prediction ability than the standard BP NN model.