Analysis And Prediction Of Deep Excavation Deformation Based On Time Series Metabolic Method

As one of the most common underground engineering,deep excavation engineering itself and the surrounding environment impact control is the most critical issue in the project management.With the improvement of environmental protection requirements,the deformation control of deep excavation has become the key problem to ensure the safety of excavation supporting structure and surrounding environment.Deformation monitoring and safety assessment of deep excavation play a key role in the construction management of deep excavation.Analysis forecast is an important part of the safety management of deformation monitoring data of deep excavation by using mathematical model or mechanics,but common mathematical models,such as the statistic model,deterministic model,the function needs to be determined and can not accurately predict the deformation of excavation development.At present,there is still lack of simple and rapid analysis and prediction method suitable for site management of excavation project.Time series analysis is a method to predict the future information when the various factors of the original system do not change greatly,considering various factors comprehensively,and using the historical information of the system itself to predict the future.Aiming at the accuracy of time series model will reduce with time of the shortcomings,the introduction of new information in a timely manner by using the method of analysis is put forward The new supersedes the old.,prediction method of deep excavation deformation time series-The new supersedes the old.method based on practical engineering application verification,provide rapid assessment methods for security management of deep excavation engineering.The main research contents and results are as follows:(1)based on time series analysis of the basic concepts and methods,a comparative analysis of the different classical models of time series(including AR(P)model,MA(q)model,ARMA(P,q)model)discusses the model checking method,parameter estimation,model identification,shortage of time sequence method.(2)a method of excavation deformation analysis and prediction based on time series metabolism method is proposed.The method of metabolism is used to optimize the time series model,and the data information is updated in real time.The analysis method is optimized,and the monitoring data of the engineering example is analyzed and modeled.Through the comparison and prediction,it is concluded that the improvement of the metabolic method is effective.(3)at last,the monitoring data of engineering case are analyzed and modeled by using the time series metabolism method described above,and the model is used to predict the monitoring data.Through this study,the feasibility and effectiveness of the metabolic model based on time series are analyzed and perfected.The example shows that the established prediction model based on time series metabolic model can more accurately predict the deformation of deep excavation,so as to ensure the construction safety of deep excavation,and shows that this method has good application value.