基于高阶矩阵函数的广义逆波束形成改进算法

发布时间：2018-02-26 18:13

本文关键词： 传声器阵列声源识别广义逆波束形成高阶矩阵函数　出处：《振动与冲击》2017年10期 　论文类型：期刊论文

【摘要】：广义逆波束形成是一种高效的声源识别定位方法,然而其计算稳健性易受随机噪声影响,阻碍了其声源识别动力学水平进一步提高。为改善广义逆波束形成声源识别方法的稳健性,基于高阶矩阵函数提出一种广义逆波束形成改进算法:定义了基于广义逆波束形成的正则化矩阵;对正则化矩阵与波束形成输出进行迭代运算;利用高阶矩阵函数对迭代求解所得广义逆波束形成输出的互谱进行优化。通过数值仿真详细分析了声源频率对波束形成矩阵函数阶次取值的影响,得到阶次的最优取值区间。最后通过数值模型和实验算例对单极子与相干声源进行定位识别,结果表明:改进算法在准确识别声源基础上能有效抑制旁瓣干扰,且具有更高的声源识别精度。
[Abstract]:Generalized inverse beamforming is an efficient method for acoustic source identification and localization, but its computational robustness is easily affected by random noise. In order to improve the robustness of the generalized inverse beamforming acoustic source recognition method, An improved algorithm of generalized inverse beamforming based on higher-order matrix function is proposed: the regularization matrix based on generalized inverse beamforming is defined, and the regularization matrix and beamforming output are iterated. The cross-spectrum of the generalized inverse beamforming output obtained by iteration is optimized by using high-order matrix function. The influence of sound source frequency on the order of beamforming matrix function is analyzed in detail by numerical simulation. The optimal value interval of order is obtained. Finally, the location recognition of monopole and coherent sound source is carried out by numerical model and experimental example. The results show that the improved algorithm can effectively suppress sidelobe interference on the basis of accurate identification of sound source. And has higher sound source recognition accuracy.
【作者单位】：重庆大学汽车工程学院;重庆大学机械传动国家重点实验室;
【基金】：国家自然科学基金(51275540) 重庆市重大应用开发计划项目(cstc2015yykf C60003)
【分类号】：TN641

【相似文献】