A new method for feature extraction of ship-radiated noise based on successive variational mode decomposition and dispersion entropy

The feature extraction of ship-radiated noise can support ship target identification and is widely applied in ship detection and identification. However, the feature extraction method of ship-radiated noise have poor recognition effect and poor separability. To address these limitations, a feature extraction method based on successive variational mode decomposition (SVMD) and dispersion entropy (DE) is proposed in this paper. This method creatively combines SVMD with DE. First, SVMD is employed to perform adaptive modal decomposition on the signal. Subsequently, for the valid modal components obtained from the decomposition, DE is utilized to characterize the irregularities and dynamic variation features of each modal component. Finally, the validity of the proposed method is confirmed using real-world ship dataset.