【摘要】 根据理论分析，提出了一种主从结构的自适应滤波算法（ＡμＬＭＳ），它利用一个从自适应滤波器，按照最小均方误差准则，自动寻找主自适应滤波器的最佳步长矢量，使滤波器的特性得到显著的提高。计算机模拟实验结果验证了其良好的性能。该主从结构自适应滤波算法的思想可拓展至其它一些自适应滤波算法中去建立其相应的主从结构更多还原

【Abstract】 Based on theoretical analysis,an adaptive filtering system with main subordinate architecture is proposed,which is named as adaptive stepsize(μ) LMS algorithm(AμLMS).According to the least mean square error criterion,the subordinate adaptive filter searchs the optimal stepsize vector of the main adaptive filter.The performance of AμLMS algorithm is much better than that of LMS algorithm,which is verified by computer simulation.The main subordinate architecture can be applied with some other adaptive filtering algorithms.更多还原