Application Research Of Regression-BP Combination Model Of Wavelet Denoising In Deformation Monitoring Of Deep Foundation Pit

With the acceleration of the urbanization process,the scale of its development is continuously expanding,and there are more and more large-scale infrastructures.However,due to the common external forces of natural factors and human activities,the safety of deep foundation pit engineering has become a constraint to society An important factor in economic development.In engineering construction,deformation monitoring runs through the construction and operation management stages of the entire project and is an important issue that cannot be ignored.At present,the methods used for deformation prediction can be roughly divided into data analysis methods based on actual measurement and calculation methods based on theory.Due to the progress of science and the need for data prediction,many prediction models have emerged based on the analysis of measured data.In practical engineering applications,by analyzing the advantages of each model,if the advantages of different prediction models can be combined to establish a comparative A good combined model can improve the accuracy of deformation prediction to a certain extent,and further guarantee the safety of deep foundation pit engineering.This article takes a community in Tai’an as an example,through the discussion and analysis of the establishment of deformation monitoring network,observation methods,monitoring accuracy,etc.,on the basis of wavelet denoising to study and establish a regression analysis model,BP neural network and regression-BP combination model,The original settlement data is analyzed and predicted,and the specific research content is as follows:(1)According to the actual observation data,the MATLAB program is used to establish and analyze the regression analysis model and the BP neural network model for the monitoring points.According to the prediction results of different models,the advantages and disadvantages of the respective models are analyzed and compared.(2)With the help of MATLAB program and EXCEL tool,through the comparison of different threshold functions,different SCALs,different decomposition layers and different wavelet basis functions,the signal-to-noise ratio and root mean square error are used as accuracy evaluation criteria,and the denoising of the example data is completed deal with.(3)Using the established regression analysis,BP neural network,and regression-BP combination model,the data obtained after wavelet denoising are predicted,and the prediction results are compared and analyzed.(4)The comparison results of the accuracy evaluation standards SSE,MSE,MAPE and RMSE found that the prediction result of the single model after wavelet denoising is better than the corresponding single model before wavelet denoising,and the prediction result of the combined model after wavelet denoising is better than wavelet The combined model before denoising and the combined model after denoising have higher prediction accuracy than the single model after denoising,and finally the wavelet denoising regression-BP combined model is more efficient and accurate for predicting and analyzing the deformation of deep foundation pits.