An Efficient Noise Elimination Method for Non-stationary and Non-linear Signals by Averaging Decomposed Components

In this paper, a moving-average method of smoothing noise based on complex exponential decomposition is applied to eliminate non-stationary signal and non-linear produced by Bouc–Wen model, which are added white Gaussian simulate the in measured signal. The uses sliding window cutting entire and/or into small segments considers that stable linear. decomposed series components via decomposition, high-energy reserved reconstruct de-noised Then, due overlap reconstructed segments, average value at same time point reconstruction regarded as data. A selected investigate performance proposed method, results show has better de-noising efficiency compared with wavelet shrinkage Savitzky–Golay filter EMD (EMD-SG) for dealing signals SNR 10 dB, 15 20 using highest signal-to-noise ratio (SNR) least root mean square error (RMSE).