摘要
为了得到一种抗干扰性能强、收敛速度快、稳态误差小的变步长LMS算法,本文分析了传统LMS算法、变步长LMS算法及其改进算法,并对现有变步长LMS算法进行分类归纳,在列出几种具有代表性算法的基础上,提出了一种改进的变步长LMS算法。新方法引入修正系数,通过当前和过去估计误差以及误差的时域均值来调节自适应算法的步长,具有收敛速度快,稳态误差小的优点,并且提高了LMS算法的抗干扰性能。文中最后给出了仿真结果,仿真结果与理论分析一致。
In order to find a variable step size LMS algorithm, which has super anti-jamming capability, faster constringency and little steady error, in this paper, a brief discussion on traditional LMS, variable step size LMS algorithm and its improved algorithm are analyzed, the variable step size LMS algorithms are discussed and classified into several typical categories, and then a modified LMS algorithm of variable step size is proposed. In the algorithm, a modified coefficient is introduced, and the step size of the new algorithm is measured by the estimation of the error vector and its time field average value. So the algorithm has the excellences of faster constringency and little steady error,and the capability of anti-jamming is improved. Finally, simulation results are also provided, which prove the theory analysis.
出处
《电子测量技术》
2007年第6期52-54,99,共4页
Electronic Measurement Technology
