An improved method for extracting the characteristics of ship radiation noise

This paper proposes an intrinsic modal function (IMF) screening and reconstruction method based on variational mode decomposition (VMD) and singular spectral entropy, in which the IMF obtained by VMD is screened by information entropy calculation, the high-energy background noise is removed from low signal-to-noise ratio (SNR) signals, and the feature information of the signals is retained to improve the SNR of ship radiated noise signals, which can improve the recognition effect of ship targets to some extent. The experimental results show that the IMF reconstructed signal after information entropy screening highlights the original features of the ship’s radiated noise signal, which can effectively improve the detection rate of targets at longer distances. This technique is crucial for directing the study and creation of passive sonar surveillance systems.