Resonance-Based Time-Frequency Manifold for Feature Extraction of Ship-Radiated Noise.

Jiaquan Yan,Haixin Sun,Hailan Chen,Naveed Junejo,En Cheng

doi:10.3390/s18040936

Abstract

In this paper, a novel time-frequency signature using resonance-based sparse signal decomposition (RSSD), phase space reconstruction (PSR), time-frequency distribution (TFD) and manifold learning is proposed for feature extraction of ship-radiated noise, which is called resonance-based time-frequency manifold (RTFM). This is suitable for analyzing signals with oscillatory, non-stationary and non-linear characteristics in a situation of serious noise pollution. Unlike the traditional methods which are sensitive to noise and just consider one side of oscillatory, non-stationary and non-linear characteristics, the proposed RTFM can provide the intact feature signature of all these characteristics in the form of a time-frequency signature by the following steps: first, RSSD is employed on the raw signal to extract the high-oscillatory component and abandon the low-oscillatory component. Second, PSR is performed on the high-oscillatory component to map the one-dimensional signal to the high-dimensional phase space. Third, TFD is employed to reveal non-stationary information in the phase space. Finally, manifold learning is applied to the TFDs to fetch the intrinsic non-linear manifold. A proportional addition of the top two RTFMs is adopted to produce the improved RTFM signature. All of the case studies are validated on real audio recordings of ship-radiated noise. Case studies of ship-radiated noise on different datasets and various degrees of noise pollution manifest the effectiveness and robustness of the proposed method.

Highlights

Underwater ship-radiated noise, in which entire spectra are widely distributed from as low as x (t) = s(t) + n(t)Sensors 2018, 18, 936; doi:10.3390/s18040936 (1)www.mdpi.com/journal/sensorsSensors 2018, 18, 936 where x (t) is the raw signal collected by the hydrophone, s(t) is the clean signal of ship-radiated noise, and n(t) denotes complex environmental noise.The techniques of feature extraction and noise mitigation are extensively applied to underwater targets such as underwater acoustic signal detection, sea-bottom exploration and marine biological monitoring [8] etc
According to their non-stationary nature, emerged time-frequency analysis techniques are much more suitable for non-stationary signals for combining the advantages of methods that provide the non-stationary information in the time domain and frequency domain, such as the short-time Fourier transform (STFT) [12,13], wavelet transform (WT) [7,14] and the Hilbert–Huang transform (HHT) [15,16]
Due to the generating mechanism of ship-radiated noise and the effect of underwater acoustic channels, a signal of ship-radiated noise has the characteristics of oscillation, non-stationary and non-linear. By considering these three characteristics simultaneously, we propose a new technique for extracting the time-frequency features of ship-radiated noise called resonance-based time-frequency manifold (RTFM)

Summary

Introduction

The oscillation nature, duffing oscillator [10] and stochastic resonance theory [11] are utilized to detect the line-spectrum of ship-radiated noise. According to their non-stationary nature, emerged time-frequency analysis techniques are much more suitable for non-stationary signals for combining the advantages of methods that provide the non-stationary information in the time domain and frequency domain, such as the short-time Fourier transform (STFT) [12,13], wavelet transform (WT) [7,14] and the Hilbert–Huang transform (HHT) [15,16]. Taking into consideration the non-linear nature of ship-radiated noise, numerous methods are employed for non-linear feature extraction, including phase space reconstruction [17,18], fractal-based approaches [19,20] and complexity measures [21], etc

Methods

Results

Conclusion