Study on the Prediction Model and Demonstration of Surface Settlement in Subway Tunnel Construction
|School||Huazhong University of Science and Technology|
|Keywords||Subway tunnel Surface settlement Neural network Time series Prediction model Sensitivity analysis|
With the metro project in China’s rapid development, surface settlement caused by subway tunnel excavation affect people’s daily lives increasingly. But because of the complexity of the soil media, surface settlement is still in the immature stage. With the widespread application of computer technology, mass of monitoring data will be sorted, classified, summarized, and combined with computer software to predict the maximum settlement and the timing of surface settlement, to provide technical support and theoretical guidance for the safety warning of the metro construction. To improve the safety of subway construction is important.In this thesis, on the basis of literature reading, combined with the background of Wuhan subway construction. Firstly, the characteristics of the underground tunnel and construction method are introduced. Factors of ground settlement are analyzed and summarized. Secondly, introduce the application of BP neural network model, through empirical analysis, to predict the maximum settlement and the factors sensitivity analysis. Finally, introduce time series modeling method, through empirical analysis, to predict the timing of settlement monitoring points. Combined with applications of the Matlab and the Eviews software, this thesis predicts surface settlement from space, time respectively. Through empirical analysis, the two made a good prediction model fitting results and can be applied to the safety warning system of subway construction.This study relates to civil engineering, computer technology, econometrics, project management, etc. Two prediction models are proposed. BP neural network theory and time series theory are applied to the subway construction, to promote the research of surface settlement by subway tunnel has important practical significance.