论文部分内容阅读
物体入水声是一种瞬态信号,其波形由击水声和若干气泡脉动组成。传统的矢量信号处理方法对此类瞬态信号的检测和测向会出现困难,尤其是在信噪比较低时检测不到入水声信号。经验模态分解是一种突出信号局部瞬态特性的非线性分析方法,将矢量传感器接收的声压、振速信息分解为不同的固有模态函数,利用文中提出的模态声强器的方位估计算法,可以实现瞬态信号的检测和测向。湖试和海试结果表明该方法能把本地干扰和入水声分解到不同的模态函数中,利用模态声强器可以在本地强干扰下有效检测到入水声信号出现的时间,并可以实现测向。
The sound of water into a body is a transient signal whose waveform consists of a stroke of sound and a few pulsations of bubbles. The traditional vector signal processing method for such transient signal detection and direction will be difficult, especially at low signal to noise ratio can not detect the incoming acoustic signal. Empirical mode decomposition is a non-linear analysis method that highlights the local transient characteristics of the signal. The acoustic pressure and velocity information received by the vector sensor are decomposed into different intrinsic mode functions. By using the modal acoustic intensity position The estimation algorithm can detect and measure the transient signal. The results of the lake test and the sea test show that the method can decompose the local interference and water into different modal functions, and the modal sound intensity detector can effectively detect the occurrence time of the acoustic signal under strong local interference and can achieve Direction finding.