基于CEEMDAN和小波包分解的闸门振动信号降噪研究OA

This paper proposes an algorithm to address noise denoising in vibration signals from gate monitoring.The proposed noise reduction algorithm combines the Complementary Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN)and wavelet packet decomposition to effectively eliminate external interference in vibration signals of sluice gates at hydropower plants.The CEEMDAN method is utilized to separate external disturbances from real signal components by decomposing the signals into several intrinsic mode functions(IMFs),each representing vibration components at different frequencies.The subsequent wavelet packet decomposition is applied to further decompose each IMF into sub-bands at varying scales and frequencies,enabling more accurate location and separation of interference components.Noise is then effectively removed from each sub-band by applying the threshold denoising technology,ensuring the generation of real signals.Experimental results showed the successful elimination of useless noise from gate vibration signals using this algorithm.The proposed algorithm offers an effective approach to improve the accuracy of gate vibration signals.