中文摘要：

针对近场声源定位问题,提出一种基于奇异值分解的稀疏重构定位方法。该方法通过奇异值分解得到信号子空间,然后在信号子空间约束l1范数求解优化问题实现声源的定位。与直接对接收信号进行稀疏重构相比,该方法通过奇异值分解降低了计算量,有效抑制了噪声。仿真结果表明,与子空间方法相比,其提高了定位的抗噪声性能和分辨率。

英文摘要：

Aiming at the problem of near-field sound sources localization,this paper proposed the method of sparse reconstruction based on the singular value decomposition（ SVD）. By the method of singular value decomposition to obtain the signal subspace,it minimized the l1 norm the sparse representation of sensor measurements,and realized the sources localization.Compared with the direct l1 norm sparse reconstruction,the method of singular value decomposition reduces the amount of calculation. Compared with subspace method,the SVD method improves the performance of resisting noise and resolution.