论文部分内容阅读
相位麦克风阵列技术是最近十年快速发展起来的一种声学实验技术,旨在精确地识别和定位出噪声源,反演出每个声源的频率、位置和幅值信息。然而对于复杂分布声源,已有的麦克风阵列后处理技术难以获得令人满意的结果,声源频率通常可以精确识别,而位置和幅值重构误差较大。为此,本文引用了一种正则化技术来寻求获得精确解,进行了相关的数值验证和实验验证。结果表明,采用正则化方法能够很好地解决反问题算法中存在的幅值误差放大现象,从而比较准确地识别和定位出声源信息。
Phase microphone array technology is a rapid development in the last decade, an acoustic experimental technique to accurately identify and locate the noise source, the inversion of each sound source frequency, position and amplitude information. However, for the complex distribution of sound sources, the existing microphone array post-processing technology is difficult to obtain satisfactory results, the sound source frequency can usually be accurately identified, and the location and amplitude reconstruction error. To this end, this paper cites a regularization technique to seek the exact solutions, and carries out the related numerical and experimental verification. The results show that using the regularization method can well solve the amplitude error amplification problem existing in the inverse problem algorithm, so as to identify and locate the sound source information more accurately.