Abstract
To improve the recognition accuracy of ship-radiated noise, a feature extraction method based on regenerated phase-shifted sinusoid-assisted empirical mode decomposition (RPSEMD), mutual information (MI), and differential symbolic entropy (DSE) is proposed in this paper. RPSEMD is an improved empirical mode decomposition (EMD) that alleviates the mode mixing problem of EMD. DSE is a new tool to quantify the complexity of nonlinear time series. It not only has high computational efficiency, but also can measure the nonlinear complexity of short time series. Firstly, the ship-radiated noise is decomposed into a series of intrinsic mode functions (IMFs) by RPSEMD, and the DSE of each IMF is calculated. Then, the MI between each IMF and the original signal is calculated; the sum of MIs is taken as the denominator; and each normalized MI (norMI) is obtained. Finally, each norMI is used as the weight coefficient to weight the corresponding DSE, and the weighted DSE (WDSE) is obtained. The WDSEs are sent into the support vector machine (SVM) classifier to classify and recognize three types of ship-radiated noise. The experimental results demonstrate that the recognition rate of the proposed method reaches 98.3333%. Consequently, the proposed WDSE method can effectively achieve the classification of ships.
Highlights
By analyzing ship-radiated noise, a slice of crucial information such as the type, speed, and tonnage of the ship can be extracted
The results showed that the energy difference between the high and low frequency is at the same level for similar ships, but there is an obvious difference for different types of ships
The results demonstrate that the weighted DSE (WDSE) value is at the same level for the same ships, but there is an obvious difference for different types of ships
Summary
By analyzing ship-radiated noise, a slice of crucial information such as the type, speed, and tonnage of the ship can be extracted. To realize the feature extraction of ship-radiated noise, scholars have proposed Fourier transform, wavelet transform and modern spectrum estimation [7]. Scholars have applied the mode decomposition algorithm and entropy to various signal feature extraction fields and achieved promising results. Chen et al [6] proposed a feature extraction method of ship-radiated noise based on hierarchical cosine similar entropy. Yang et al [26] used EEMD to decompose the ship-radiated noise, and the energy difference between the high and low frequency was extracted and analyzed. A feature extraction method of ship-radiated noise based on RPSEMD, mutual information (MI), and DSE is proposed. The feature vector WDSE is put into SVM for classification
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
