Application Of Grey Time Series Combination Model Based On Denoising Method In Deformation Monitoring

The times are developing,society is progressing,and the pace of urbanization is accelerating.The emergence of high-rise buildings and the construction of bridges and tunnels require a large-scale construction team and more mature construction techniques.In addition,in order to ensure people’s life safety and property safety,construction safety personnel are also required to be able to grasp the deformation of the building in real time,and take relevant measures to ensure the safety status of the building.So as not to cause unnecessary threats to life and property.In the safety monitoring of buildings,people realized that building deformation is only a means.By analyzing the actual data that has been measured,a mathematical model can be established that can accurately predict the deformation of the building.It is the ultimate significance to provide scientific and accurate prediction for the smooth progress of the construction.This paper takes the deformation monitoring data of Taian Yihua Shopping Center as the specific background,and discusses the wavelet denoising,singular spectrum analysis,gray prediction model and time series model.Pointing out the limitations of a single predictive model,try to replace the single predictive model with a new combined model,and then compare the fit and prediction accuracy of the two.Select a more accurate prediction model.On the other hand,the method of wavelet analysis and singular spectrum analysis is used to pre-process the original observation data,and select a better denoising method by comparison.The prediction accuracy and the degree of fitting of the prediction model before denoising and the prediction model after denoising are compared.Choose the optimal predictive model.The specific research process is as follows:1)Through reviewing,analyzing and summarizing a large number of papers,books and other literatures,the paper systematically combs the techniques related to deformation monitoring at home and abroad,and details the related theories of wavelet denoising,singular spectrum analysis,gray model and time series.The modeling mechanism analyzes the shortcomings of existing research.2)Combined with the settlement data of Taian Yihua Shopping Center,the original data was denoised by wavelet analysis and singular spectrum analysis.By comparing the RMSE and PSNR data obtained from many experiments,it is found that the wavelet denoising method has a good denoising effect.3)Based on the deformation monitoring data of Taian Yihua Shopping Center.the original data was modeled using gray model,time series model and grey-time series combination model.Through the comparison of prediction accuracy and prediction effect.Choose a grey-time series combination model with a relatively better prediction effect.4)In order to find the best predictive model,wavelet denoising with good denoising effect is obtained through the previous exploration experiments.First,the original data is denoised,and the denoised data is modeled by combining gray model,time series model and gray-time series model.The prediction accuracy and prediction effect of the single model before and after denoising and the combined model are compared.A more complete prediction model is obtained—a gray time series model based on wavelet denoising.