High-G Calibration Denoising Method for High-G MEMS Accelerometer Based on EMD and Wavelet Threshold.

Area of Science:

• Engineering
• Signal Processing
• Geophysics

Background:

• High-G MEMS accelerometers are crucial for natural disaster monitoring.
• Existing accelerometers face performance limitations due to noise.
• Improving signal fidelity is essential for accurate data acquisition.

Purpose of the Study:

• To develop and validate an advanced denoising method for High-G MEMS accelerometers.
• To enhance the performance and reliability of accelerometers in critical applications.
• To effectively suppress noise while preserving essential signal characteristics.

Main Methods:

• Empirical Mode Decomposition (EMD) to decompose accelerometer output into Intrinsic Mode Functions (IMFs).
• Wavelet threshold denoising applied to high-frequency IMFs.
• Continuous Mean Square Error rule for energy cut-off point determination.
• Reconstruction by superposing denoised high-frequency IMFs with low-frequency IMFs.

Main Results:

• The proposed method effectively integrates EMD and wavelet thresholding advantages.
• Noise in the original signal was reduced by 96%.
• Impact peak and vibration characteristics were preserved at 0.003% and 0.135% of the original signal, respectively.

Conclusions:

• The combined EMD and wavelet threshold method offers superior denoising for High-G MEMS accelerometers.
• This technique maximizes the retention of useful signal components.
• The enhanced performance makes accelerometers more reliable for natural disaster monitoring and other demanding fields.