基于高阶统计量自适应盲源分离算法刘涵,刘丁,刘筱琰.pdfVIP

　西安理工大学学报 (2002) . 18 . 2　 Journal of X i’an U niversity of T echno logy V o l N o 113 　　文章编号: (2002) 基于高阶统计量的自适应盲源分离算法 刘　涵, 刘　丁, 刘筱琰 (西安理工大学 自动化与信息工程学院, 陕西 西安 710048) 摘要: 提出了一种新的自适应盲源分离算法。在无噪音实时两源两传感器的情况下, 一旦观 测信号被白化, 只需要辨识一个特定的旋转矩阵就可以完成盲源分离, 并给出了能表征该旋 转矩阵的角的自适应估计器。仿真结果表明, 当满足源峭度和不为零的条件时, 这种方法是 一种稳定的和有效的分离算法。 关键词: 盲源分离; 概率密度函数; 高阶统计量 中图分类号: TN 912. 3　　　文献标识码: A Blind Sources Separa tion Using H igher Order Sta tistic , , L IU H an L IU D ing L IU X iao yan ( , 710048, ) X i’an U niversity of T echno logy X i’an Ch ina Abstract: A new learn ing algo rithm is developed fo r b lind separation of independen t sou rce . signals from their linear m ix tu res In the no iseless real m ix tu re tw o sou rce tw o sen so r scenario , once the ob servation s are w h itened (deco rrelated and no rm alized) , on ly a given ro tation m atrix rem ain s to be iden tified in o rder to ach ieve the sou rce separation. In th is paper , an adap ter estim ato r of the angle that characterizes such a ro tation is derived. It show s that estim ato r converges to a stab le valid separation so lu tion w ith the on ly condition that the sum of sou rce ku rto sis be distinct from zero. Sim u lation dem on strate the validity of the algo rithm. Key words: b lind sou rce separation; p robab ility den sity function; h igher o rder statistic