中文摘要：

仿射投影算法（APA）重复利用数据，可提高算法的收敛速度。针对现有盲源分离（BSS）收敛慢问题，以BSS的独立分量分析（ICA）为基础，结合APA思想，设计出BSS的APA—ME、APA—MMI、APA-EASI新算法。在这些新算法中，输出向量数据被重复利用，向量式数据转变成矩阵式数据，从而加快了BSS的收敛速度。仿真结果表明，APA—ICA类的BSS算法是有效的。

英文摘要：

Using data in a repeated mode can improve the convergence speed of the affine projection algorithm （APA）. Ai- ming at the problem of slow convergence in the existing blind source separation （BSS） , based on the independent component analysis （ICA） for BSS, this paper designed new APA-ME, APA-MMI and APA-EASI algorithms for BSS by using the idea of APA. In these new algorithms, output vector data was utilized in a repeated fashion, and the vector data was thus converted into matrix data. The convergence rate of BSS was accelerated. Simulation results show that the effectiveness and applicability of APA-ICA class BSS algorithm.