一种改进的变步长LMS自适应滤波算法  

An improved variable step size LMS adaptive filtering algorithm

作　　者：张继荣 张天 ZHANG Jirong;ZHANG Tian(School of Communications and Information Engineering,Xi'an University of Posts and Telecommunications,Xi'an 710121,China)

机构地区：[1]西安邮电大学通信与信息工程学院,陕西西安710121

出　　处：《西安邮电大学学报》2021年第1期7-12,共6页Journal of Xi’an University of Posts and Telecommunications

基　　金：国家电网公司系统科学技术项目(5217C0160002)。

摘　　要：针对自适应滤波领域的最小均方(Least Mean Square,LMS)算法无法权衡稳态误差和收敛速度这一矛盾,提出了一种改进的变步长LMS自适应滤波算法。该算法在基于对数函数的变步长LMS算法的基础上,建立了一种新的步长参数与误差的关系模型。仿真结果表明,提出算法与已有算法相比,能够达到更高的收敛精度及更快的收敛速度,在系统不发生时变时,收敛精度分别提高了5 dB和3 dB,当系统发生时变后,收敛精度分别提高了4 dB和2 dB,不论系统是否发生时变,收敛速度都更快。For the contradiction that least mean square(LMS)algorithm cannot balance the steady-state error and the convergence speed,a variable step size LMS adaptive filtering algorithm is proposed.On the basis of the logarithmic function based variable step size LMS algorithm,a new relationship model between step size parameters and error is established.The simulation results show that compared with the algorithms in the literature,higher convergence precision and faster convergence speed can be achieved.More specifically,the convergence precision is improved by 5 dB and 3 dB respectively when the system is not time-varying,and the convergence precision of the system is improved by 4 dB and2 dB respectively under the system varying condition,The convergence rate is faster whether the system is time-varying or not.

关 键 词：LMS算法 自适应滤波器 收敛速度 跟踪性能 收敛精度 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP301.6[自动化与计算机技术—控制科学与工程]