Improving purity of blasting vibration signals using advanced Empirical Mode Decomposition and Wavelet packet technique

The study proposes a denoising method based on Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) and Wavelet packet threshold denoising (WPTD) to reduce the noise of blasting vibration signals. CEEMDAN is first employed to decompose noisy blasting vibration signals into different intrinsic mode functions (IMFs), i.e., different frequencies of subcomponents. By using cor-relation coefficients, all of IMFs are automatically categorized as the dominant IMFs with higher portions of vibration components and the secondary IMFs with less vibration contents. Then, with further decom-posing two groups of IMFs into different scales of Wavelet packet coefficient series, WPTD based on the fixed-threshold technique and the soft threshold function method is applied to minimize the level of noise in two groups of IMFs. Two groups of denoised IMFs are finally reconstructed as the pure blasting vibration waveform. Numerical waveforms with different types and intensities of noise and measured blasting vibration signals with different main frequency bands are used to verify the accuracy and capa-bility of the proposed method. The numerical and field results show that the proposed method can not only effectively reduce noise components, but also well remain components of blasting vibrations.(c) 2022 Elsevier Ltd. All rights reserved.

Automated summary

The study proposes a denoising method based on CEEMDAN and WPTD to reduce the noise of blasting vibration signals. The method decomposes the signals and applies wavelet packet threshold denoising to effectively reduce noise and preserve vibration components.