中文摘要：

基于到达时间差（Time difference of arrival，TDOA）估计的方法是声源波达方向（Direction of arrival，DOA）估计中的一类重要方法。其中由TDOA到DOA的映射是该类方法的关键步骤。本文提出了一种基于多核聚类最小二乘支持向量回归（Least-squares support vector regression，LS-SVR）的TDOA-DOA映射方法，并且分析了其稀疏化处理后的性能。为了提高混响噪声环境下的TDOA-DOA映射性能，本文还给出了一种基于归一化中值滤波的TDOA估计离群值消除方法。仿真结果表明，本文提出的方法要优于现有的最小二乘方法以及单核LS-SVR方法。

英文摘要：

In sound source direction of arrival （DOA） estimation, one of the typical methods is based on the time difference of arrival （TDOA）. For the TDOA-based sound source DOA estimation, the TDOA- DOA mapping is a crucial step. Here, we propose a TDOA-DOA mapping approach based on the multi- kernel least-squares support vector regression （LS-SVR）, and also analyze its performance with sparsifi- cation. In addition, we present an outlier detection method based on the normalized median filtering to post-process the TDOA estimation for improving the performance of TDOA-DOA mapping in noisy re- verberant environments. Simulation results show that the proposed method is superior to its counter- parts, such as LS and single-kernel LS-SVR methods.