A New Ship-Radiated Noise Feature Extraction Technique Based on Variational Mode Decomposition and Fluctuation-Based...

Area of Science:

• Marine acoustics
• Signal processing
• Machine learning

Background:

• Underwater acoustic channels present challenges for ship-radiated noise signal feature extraction.
• Existing methods struggle with the complexity of the sea environment.

Purpose of the Study:

• To develop a novel and precise feature extraction method for ship-radiated noise signals.
• To improve the classification accuracy of underwater acoustic signals.

Main Methods:

• Variational Mode Decomposition (VMD) to decompose signals into intrinsic mode functions (IMFs).
• Fluctuation-based Dispersion Entropy (FDE) applied to IMFs to identify the most representative component (SIMF).
• Self-Organizing Feature Map (SOM) for classifying the extracted features.

Main Results:

• The proposed VMD-FDE-SOM method demonstrated higher precision in ship-radiated noise feature extraction.
• Successful application to real-world ship-radiated noise signal datasets.
• Effective categorization of signals using the developed characteristic vectors.

Conclusions:

• The integrated VMD-FDE-SOM approach offers a robust solution for ship-radiated noise analysis.
• This method overcomes limitations of traditional feature extraction in complex marine environments.
• The findings contribute to improved underwater acoustic signal processing and identification.