Analysis And Application Of Multiple Regression Optimization Model In Subway Surface Settlement Deformation

Today in the 21 st century,China has made substantial breakthroughs in many fields.In terms of rail transit,subway has become a household name.People's demand for the convenience of modern transportation is increasing,and the construction of modern transportation by the state is a top priority.However,as the main tools of underground transportation in the new era,subway not only brings people the fast speed and convenience of transportation,but also lays hidden dangers for the surrounding buildings,the existing roads and the engineering construction.For these hidden dangers,the first is the road surface and tunnel collapse caused by excessive settlement deformation in the construction process of the subway,and the second is the deformation of the surface and surrounding buildings that the subway passes through after its completion.No matter the subway under construction or the subway that has been built,it will be a very serious safety problem if all kinds of deformation problems cannot be controlled in time.In order to avoid such problems,it is particularly important to conduct a comprehensive deformation monitoring process for all subway projects under construction or completed.At present,the analysis methods of subway tunnel deformation mainly include: regression analysis,grey theory model,wavelet analysis theory,BP neural network and time series model and so on.This paper mainly takes the multivariate regression model as the basic model,through the optimization and screening of independent variable parameters,de-noising of The original data,and the optimization of multiple regression coefficients,finally,the overall optimization of multiple regression model is realized.In terms of data de-noising,the denoising precision of the resistance difference kalman filter model is compared with that of the wavelet decomposition and reconstruction transformation soft threshold denoising model,so as to obtain the data with the highest denoising accuracy as the basis for optimizing the multiple regression coefficient and establishing the optimization model.Through the comparison before and after the optimization of multiple regression model,the advantages of the optimized multiple regression model are analyzed,the optimization model is applied to the subsidence and deformation engineering of the surface of the subway construction area,and the analysis and prediction are made based on the actual construction situation,so as to maximize the benefits of the application of the optimization model.This paper first introduces the purpose and significance of metro tunnel deformation monitoring,deformation monitoring technology and deformation monitoring data processing method.Secondly,the theory and application of multiple regression models in the regression model are introduced.In the theory of wavelet analysis,several commonly used wavelet functions are described,and the characteristics of these wavelet functions are analyzed.The function of wavelet de-noising is realized by MATLAB software.Finally,through the example of surface subsidence project in metro construction area,the optimization of multiple regression model is realized by the cross application of multiple methods,and the advantages of the optimized multiple regression model are summarized.