常规声压与声矢量信号非整数维谱性能比较研究

为解决高阶谱算法复杂、计算量大和声压信号有限的抗干扰能力问题,提出了声矢量信号非整数维谱分析方法.利用可抑制高斯和对称分布噪声的高阶累积量非整数维谱对目标辐射噪声声压信号和声矢量信号进行了特征分析.分别采用功率谱图和三维动态谱图方法对常规声压和声矢量信号非整数维谱性能进行了直观比对.为获得定量分析结果,分别对不同背景噪声环境条件下,不同输入信噪比的常规声压与声矢量信号非整数维谱的轴频PBR及PD进行了详细计算.结果表明,声矢量信号非整数维谱特征提取与轴频检测能力优于常规声压信号,为高阶统计量应用于声矢量信号处理提供了一条途径.

Comparing sound pressure signals and acoustic vector signals by nonintegral dimension spectrum

To resolve issues of computational complexity with higher-order spectra, as well as limit antijamming capabilities inherit with sound pressure signals, a nonintegral dimensional spectrum is proposed for analyzing sound pressure signals. The characteristics of pressure and vector signals of target radiated noise are analyzed by a nonintegral dimensional spectrum that can theoretically restrict high order cumulants of Gausslan and symmetrical distribution noise. Intuitionistic comparisons are made with respect to nonintegral dimensional spectrum performance of sound pressure and acoustic vector signals by using power spectrum and 3D dynamic spectrum analysis respectively. To get quantitative analysis results, the PBR and PD of propeller shaft frequency with different sound pressures and acoustic vector signals were calculated under various noise backgrounds and different SNR. Simulation suggests that feature extraction and detection of propeller shaft frequency through acoustic vector signal processing based on nonintegral dimensional spectrum is more effecive than when using conventional pressure signals. Using a nonintegral dimensional spectrum with higher-order statistics provides a new way for lower frequency acoustic vector signal processing.