Denoising of blasting vibration signals based on CEEMDAN-ICA algorithm

Area of Science:

• Geophysics
• Signal Processing
• Vibration Analysis

Background:

• Blasting vibration signals are often contaminated by noise from construction environments and equipment.
• Effective signal preprocessing is crucial for accurate analysis of blasting vibrations.

Purpose of the Study:

• To propose and evaluate a new denoising method for blasting vibration signals.
• To effectively alleviate noise components and retain effective signal information.

Main Methods:

• Utilized CEEMDAN (Complementary Ensemble Empirical Mode Decomposition) to decompose signals into intrinsic mode functions (IMFs).
• Applied ICA (Independent Component Analysis) to estimate and extract independent components.
• Reconstructed the signal using arrangement entropy of ICA components for denoising.

Main Results:

• The proposed CEEMDAN-ICA method demonstrated superior denoising performance compared to traditional algorithms.
• The algorithm effectively reduced noise while preserving the integrity of the original blasting vibration signals.
• Simulations confirmed the feasibility and efficiency of the CEEMDAN-ICA approach.

Conclusions:

• The CEEMDAN-ICA method offers a significant advancement in denoising blasting vibration signals.
• This technique provides a reliable new approach for subsequent signal analysis in geophysics and engineering.
• Accurate denoising is essential for understanding complex blasting dynamics.