Robust Passive Underwater Acoustic Detection Method for Propeller

Yongxing Song,Jingting Liu,Linlin Cao,Ning Chu,Dazhuan Wu
DOI: https://doi.org/10.1016/j.apacoust.2018.12.024
IF: 3.614
2019-01-01
Applied Acoustics
Abstract:Characteristic amplitude modulated radiated noise signals are generated as the propeller operates underwater. Generally, passive underwater acoustic detection is accomplished by processing the radiated noise of a marine's propeller. This study proposes a robust passive underwater acoustic detection method of a propeller by combining cyclostationary mechanism and principal component analysis (PCA). The proposed method can extract the characteristic frequency information of a propeller accurately. The effectiveness of the proposed method is verified according to simulated signals and real application cases. Both monocomponent and multicomponent modulated simulated signals are used to test the performance of the proposed method. The proposed method is then applied to the acoustic signals of propeller experiments and merchant ship propeller. Finally, the superiority of the proposed method is demonstrated by comparing the application results with the kurtogram and cyclostationary analysis method, during the extraction of the characteristic frequency of the propeller under low signal-to-noise ratio (SNR). The innovation of the proposed method not only realizes marine propeller detection from propeller acoustic signals but also has remarkable performance under low SNR. (C) 2018 Elsevier Ltd. All rights reserved.
What problem does this paper attempt to address?