A coarse-to-fine two-stage adaptive denoising method for ship signal

The mode decomposition-based denoising method serves as an effective technique, and has found extensive application in denoising for ship signal. However, existing denoising methods of this type suffer from problems such as excessive denoising or incomplete denoising. To tackle these problems, this paper proposes a coarse-to-fine two-stage adaptive denoising method for ship signal. The first stage performs coarse denoising on input signal to preserve signal components to the maximum extent, while the second stage conducts fine denoising to further remove residual noise. Specifically, the input signal is first decomposed via complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN). Then, intrinsic mode functions (IMFs) are classified using the fuzzy diversity entropy-based maximum gap method and reconstructed to acquire the first stage denoised signal. Next, a second CEEMDAN is performed, and IMFs are classified according to the adaptive correlation coefficient and permutation entropy joint criterion and noise IMFs are eliminated. Finally, noisy IMFs are processed via wavelet threshold, and signal IMFs are reconstructed to acquire the final denoised signal. Simulation experiments verify the effectiveness of the proposed method. Actual experiments based on real-world ship signals further demonstrate the proposed method achieves superior denoising performance over other methods.