论文部分内容阅读
水下爆炸声源激发的声信号包括冲击波和气泡脉动,气泡脉动严重干扰冲击波的传播特性。当气泡脉动的幅度小于冲击波的幅度时,通过对此类混合爆炸信号进行复倒谱分析,提出了基于卷积型的爆炸信号模型的同态解卷积气泡脉动去除方法,其中,针对同态滤波系统中通常采用的“梳状”滤波器会在倒谱的尖峰处产生不连续点的问题,采用了对尖峰附近的采样点进行多项式函数拟合,并根据拟合函数对尖峰采样点插值的改进型滤波方法。数值仿真和实验信号的处理及分析表明,应用该方法可以明显消除爆炸信号自相关曲线中由气泡脉动造成的对称次尖峰,并在信号的时频分布图中再现冲击波的简正波特性。这为进一步利用爆炸声源研究海洋信道的声传递函数和反演海洋环境参数提供了技术途径。
Acoustic signals excited by underwater explosion sources include shock waves and bubble pulsations, which seriously interfere with the propagation characteristics of the shock wave. When the amplitude of the bubble pulse is smaller than the amplitude of the shock wave, the convolution deconvolution bubble pulse removal method based on the convolution-type explosion signal model is proposed by the complex cepstrum analysis of the mixed explosion signal, in which, The “comb” filter commonly used in filtering systems creates the problem of discontinuities at the cepstral spikes. Polynomial function fitting is applied to the sampling points near the spikes and the spikes are sampled according to the fitting function Improved interpolation method for point interpolation. Numerical simulation and experimental signal processing and analysis show that the application of this method can obviously eliminate the symmetrical sub-peak caused by the bubble pulsation in the autocorrelation curve of the explosion signal and reproduce the normal wave characteristics of the shock wave in the time-frequency distribution of the signal. This provides a technical way to further study the acoustic transfer function of oceanic channel and the inversion of marine environmental parameters by using the explored sound source.