A denoising method for aeroengine gas path electrostatic signal of low signal-to-noise ratio based on IMFs optimized reconstruction and wavelet threshold

Electrostatic monitoring technology for aero-engines has demonstrated considerable capability in early fault warning. However, raw electrostatic signals often contain significant noise and exhibit low signal-to-noise ratios, making denoising essential to improve the accuracy of fault-related information extraction. To address the issue of coupled noise in electrostatic signals, this study introduces methodologies based on Improved Complete Ensemble EMD with Adaptive Noise (ICEEMDAN), Autocorrelation Function (ACF), and Wavelet Soft-thresholding (WTD). We investigate the criteria and principles for screening intrinsic mode functions (IMFs) and propose a joint denoising algorithm, along with its specific procedure, based on IMF-optimized reconstruction and wavelet thresholding. The proposed method is validated using both simulated signals and actual electrostatic signals collected from a micro-turbojet engine test. Comparisons with other denoising techniques are conducted. Simulation results indicate that the proposed method improves denoising performance in terms of signal-to-noise ratio (SNR), mean square error (MSE), and normalized cross-correlation (NCC). Test results further demonstrate that the method effectively suppresses random noise and power frequency interference while preserving useful abnormal particle signals more effectively.