关键词: 混合矩阵估计, 人工蜂群算法, 欠定盲分离, Levy飞行

Abstract: To solve the problem that the estimation accuracy of the mixing matrix in the underdetermined blind separation is not high, this paper proposes an improved Artificial Bee Colony（ABC） clustering algorithm. From the linear clustering characteristics of the observed signals and the diversity of bee colony, this paper improves the search strategy of the employed bee in order to speed up the convergence rate of the algorithm. At the same time, the local search method based on Levy flight is introduced to search the neighborhood of the current optimal solution, which can improve the local development ability of ABC algorithm. The simulation results show that the proposed method has higher estimation accuracy under the condition that the number of sources is large.

Key words: mixing matrix estimation, Artificial Bee Colony（ABC） algorithm, Underdetermined Blind Source Separation（UBSS）, Levy flight