A Novel Denoising Method for Ship-Radiated Noise

Ship-radiated noise (SN) is one of the most critical signals in complex marine environment; however, it inevitably contaminated by environment’s as well from other equipment. Thus, feature extraction and identification SN becomes very arduous. This paper proposes a denoising method for based on successive variational mode decomposition (SVMD), dual-threshold analysis fuzzy dispersion entropy (FuDE) wavelet packet (WPD), termed SVMD-FuDE-WPD. First, SVMD adaptively decomposes into certain intrinsic functions (IMFs), which can solve parameter selection problem (VMD) suppress mixing empirical (EMD). After that, FuDE-based used to accurately classify IMFs signal IMFs, noise–signal IMFs. Finally, denoised could be obtained reconstructing that were using WPD. The classical simulation experiments demonstrate effectiveness proposed method, performs better than four existing methods. And measured show attractor trajectories are smoother more regular, verifies method.