Denoising method for gross errors and random errors of monitoring displacement for high rock slope

Abstract

Data denoising is an important issue for data processing. The gross errors in a nonlinear time series are detected by using the three-standard-deviation rule (3-σ rule) and by reconstructing the time series by a first-order Lagrange interpolation method. Then the reconstructed time series is used to denoise the random errors by a discrete stationary wavelet transform (DSWT) method. Finally, the present data denoising method is applied to the error analysis of the slope displacement monitoring data collected at the Jinping I Hydropower Station. Computed results show that the data denoising results can be improved through removal of the gross errors and repair of the time series followed by application of wavelet transforms to denoise the random errors. © 2014 Springer Science+Business Media New York.