基于平面聚类势函数法的欠定混合信号盲分离

Underdetermined blind sources separation using the potential function method based on plane clustering

针对不充分稀疏欠定混合信号盲分离,提出了一种基于超平面聚类的势函数法来估计源信号个数和混合矩阵。该方法在源信号个数未知的情况下,利用聚类平面法线向量构成势函数,通过估计势函数的局部最大值来估计聚类平面的法线向量,然后再通过估计聚类平面的交线来实现混合矩阵的估计。为了提高算法对异常值的鲁棒性,不直接估计势函数的局部最大值,而是采用聚类算法来估计势函数的局部最大值。计算机仿真试验证实了该算法的有效性及其较好的性能。

A potential function method is proposed to estimate the number of sources and the mixing matrix for the problem of underdetennined blind separation of insufficiency sparse sources based on the hyperplane clustering algorithm. When the number of sources is unknown, the normal vectors of the concentration hyperplanes can be obtained by estimating the local maxim of the potential function, and then the mixing matrix can be estimated by finding the intersection of the concentration hyperplanes. In order to increase the robustness to the outliers, the clustering algorithm is exploited to estimate the local maxim of the potential function instead of directly estimating the local maxim of the potential function. The simulation results show the validity and high performance of the algorithm.