摘要
RLS算法具有收敛速度快,且对自相关矩阵特征值的分散性不敏感等优势,得到了广泛的研究与应用。本文重点研究了RLS算法,给出了它的具体推导过程和基本设计原理,并对其优良性能进行了简单的理论分析,然后将其应用于系统辨识,计算机仿真结果表明RLS算法比LMS算法具有更好的系统辨识效果。
RLS algorithm has the advantages of fast convergence speed and also the insensitive dispersion about the autocorrelation matrix eigenvalues make it received extensive research and application. This paper focuses on the RLS algorithm and presents its specific deducing process and basic design principles. The theoretical analysis shows the excellent performance of RLS algorithm,and then applied it to the system identification,computer simulation results show that the RLS algorithm effects better system identification than the LMS algorithm.
出处
《中山大学研究生学刊(自然科学与医学版)》
2013年第1期80-89,共10页
Journal of the Graduates Sun YAT-SEN University(Natural Sciences.Medicine)
关键词
自适应滤波
RLS算法
LMS算法
系统辨识
Adaptive Filtering
Recursive Least Square Algorithm
Least Mean Square Algorithm
System identification