Decomposition of rock micro-fracture signals based on a singular value empirical mode decomposition algorithm

Area of Science:

• Geophysics
• Mining Engineering
• Signal Processing

Background:

• Rock burst early warning systems primarily rely on microseismic monitoring.
• Rock bursts are preceded by micro-fracture events that generate detectable seismic waves.
• Existing methods struggle with low signal-to-noise ratios in microseismic data.

Purpose of the Study:

• To develop an advanced signal decomposition technique for rock burst early warning.
• To improve the precision of identifying characteristic rock micro-fracture signals.
• To enhance the reliability of pre-warning systems for rock fractures.

Main Methods:

• Applied singular value Empirical Mode Decomposition (EMD) for signal characteristic decomposition.
• Developed an improved EMD algorithm to enhance decomposition precision.
• Validated the algorithm through laboratory simulations and real-world hydro-power station data.

Main Results:

• The improved EMD algorithm effectively decomposed characteristic rock micro-fracture signals from mixed microseismic data.
• The method demonstrated superior performance in retaining local signal characteristics compared to wavelet decomposition and standard EMD.
• The algorithm achieved better decomposition precision, enabling effective separation of micro-earthquake signals.

Conclusions:

• The improved EMD algorithm provides a robust basis for identifying abnormal microseismic signals indicative of rock micro-fractures.
• This technology significantly enhances the capability for accurate rock fracture pre-warning.
• The study highlights the practical significance of advanced signal processing in rock burst early warning.