摘要
DEMON分析是舰船噪声目标识别的重要分析手段之一。通过DEMON分析可以获得诸如舰船螺旋桨转速、螺旋桨叶片甚至舰船的车数等不变的舰船物理特征。在过去的DEMON分析中,通常采用经典的周期图方法。本文采用了高阶谱分析、小波的子频带分析以及基于互相关函数矩阵奇异值分解等方法,对舰船噪声信号进行了分析。从对大量的舰船噪声信号的分析来看,在大多数情况下现代谱分析技术具有优势,但并不是所有情况下其性能均优于周期图方法。
DEMON analysis is one of the most important means for underwater target recognition. Many invariable physical features can be obtained with DEMON analysis such as revolution rate of propeller, the propeller vane, and the number of propellers. Traditionally, periodogram is used. This paper uses the high order spectrum, wavelet subband analysis, and singular value decomposition of correlation function matrix. The results show that modern spectrum analysis is more effective in most cases. However, sometimes it is inferior to the periodogram method.
出处
《声学技术》
EI
CSCD
北大核心
2006年第1期71-74,共4页
Technical Acoustics
