Wavelet denoising can remove outliers in the original data when processing data,which makes the original data sequence smoother.In this paper,a combined optimization model based on wavelet threshold denoising and gray model is used to predict the trend of building settlement and deformation.The traditional GM(1,1)model,the gray Verhulst model and the metabolic GM(1,1)model were established using the first 10 periods of observation data after denoising in an engineering project,and the prediction effects of different models were comprehensively analyzed.The results show that the metabolic GM(1,1)model based on wavelet threshold denoising has the highest prediction accuracy,the mean square error is 0.0449 mm,and the model accuracy verification lev-el is I.Therefore,the metabolic GM(1,1)combination of wavelet threshold denoising can make more scientific and accurate predic-tions on the trend of building settlement and deformation.
维普
万方数据