An ICEEMDAN-WPD based denoising method for MS signals and its engineering application

Abstract

ABSTRACT The complex installation environments for microseismic (MS) monitoring systems generate significant noise in MS signals, hindering accurate event localization. This study proposes a noise reduction method combining improved complete ensemble empirical mode decomposition with adaptive noise and wavelet packet decomposition (ICEEMDAN-WPD). The signal is initially processed with ICEEMDAN to obtain intrinsic modal functions (IMFs), which are components ordered from high to low frequencies. The correlation between the original signal and each IMF is evaluated and low-correlation IMFs are eliminated for initial noise reduction. Additionally, the WPD method further decomposes and filters the highly correlated IMFs for secondary noise reduction, enhancing the noise reduction effect. Ultimately, the signal formed by reconstructing the secondary noise reduction IMF is the complete ICEEMDAN-WPD method filtered signal. The simulation results indicate that the ICEEMDAN-WPD method enhances the signal-to-noise ratio (SNR) and preserves the original signal characteristics to the greatest extent, demonstrating a significant noise reduction capability. In particular, the application of this method to the noise reduction process of MS signals from Shuangjiangkou hydropower station has led to a significant improvement in the accuracy of subsequent MS events localization. The research results can provide reference for similar signal noise reduction processing.