Wavelet Denoising Quality Evaluation Method And Its Application In Deformation Monitoring

Wavelet Analysis localization ability with good frequency has been widely used in signal processing.Currently,wavelet de-noising has been widely used in deformation monitoring data process,geophysical inversion and so on.However,real data cannot be obtained in actual engineering applications,and it is difficult to evaluate the effect of wavelet denoising.Therefore,it is very difficult to select the best scale.There are still problems in the quality evaluation of wavelet denoising.First,although the existing traditional indicators and evaluation methods can determine the best decomposed scale to a certain extent,the obtained results are sometimes not ideal.The second is how to extract feature information from different types of measurement data and perform accurate deformation monitoring and forecasting.This article focuses on the above issues and the specific research content is as follows:(1)In view of the limitations of the traditional single-indicator and common wavelet denoising methods,the problems of wavelet denoising quality evaluation methods were systematically analyzed.A large number of simulation experiments were performed to verify the evaluation capabilities and applicability of various evaluation methods.Analyze the advantages of the traditional single index and the existing evaluation methods,combine the entropy weight method and the coefficient of variation determination method,and proposed an improved wavelet denoising evaluation method.(2)Designing simulation experiments,comparing traditional single indicators and existing evaluation methods with proposed combined weighting methods,The real RMS error was calculated based on the prior information of the clean signal.The simulation results shows that the combination weighting method can obtain the minimum value at the same decomposition scale as the real RMS error.It is proved that the improved combination weighting method can effectively determine the best decomposition scale.(3)The optimal decomposed scales are calculated using the combined weighting method for metro settlement data and compared with the existing methods.Using AIC and BIC information criteria to prove the reliability of the results,the results showed that the combination weighting method can obtain the correct optimal decomposition scale.The ARMA model was established based on the time series model to perform deformation monitoring and forecasting and achieved higher prediction accuracy.(4)For the problem of complex information of different deformation data features,the advantages of wavelet multiresolution analysis are used to extract periodic and trend items of deformation data.And then established a regression analysis model based on wavelet multiresolution analysis and compared it with a linear regression analysis model,and the prediction accuracy has been significantly improved.