A Novel Denoising Method for Mixed Noises Based on Variational Mode Decomposition and Energy Variation Ratio Function

Signal denoising is crucial in industry. However, it often faces frustration when the actual signal is non-stationary and nonlinear. To remove the noise effectively, this paper proposes a new denoising method. This paper first decomposes the original signal into the multiples limited bandwidth intrinsic mode functions (IMFs) by variational mode decomposition (VMD). The principle of the denoising method is to use partial reconstruction, with the relevant IMFs corresponding to the essential structures of the signal. Second, the energy variation ratio (EVR) function is proposed to distinguish between the effective IMFs and the non-effective IMFs. Third, retain the valuable components and remove the useless ones. Last, the denoising signal is obtained by reconstructing the useful IMFs. On simulated and real signals, the denoised results verify the effectiveness and practicability of the proposed denoising scheme for removing mixed noises and demonstrate its superiority over the other denoising approaches reported in this paper