摘要
为了进一步减小收敛速度与稳态误差之间的矛盾,提高自适应滤波算法的性能,文中提出了一种新的凸组合自适应滤波算法.该算法将变步长归一化最小均方(NVS-NLMS)算法和变步长仿射投影(NPVSS-APA)算法进行组合,并将脉冲噪声指标加入到传统混合参数中,在充分利用两种算法性能优势的同时,有效解决了传统组合算法对脉冲噪声敏感的问题.在系统辨识和声学回声消除环境下的仿真结果表明,与NVS-NLMS算法、NPVSS-APA算法以及凸组合APSA算法相比,文中算法具有较快的收敛速度和较低的稳态误差.
In order to further reduce the contradiction between convergence speed and steady-state error and improve the performance of the adaptive filtering algorithm,a new convex combination adaptive filter algorithm is proposed.The algorithm combines the variable step size normalized least mean square(NVS-NLMS)algorithm and the variable step size affine projection(NPVSS-APA)algorithm,and adds the impulse noise index to the traditional mixing parameters,while making full use of the performance advantages of the two algorithms,it effectively solves the problem of traditional combined algorithms that are sensitive to impulse noise.Simulation results in system identification and acoustic echo cancellation environment show that the proposed algorithm has faster convergence speed and lower steady-state error than NVS-NLMS algorithm,NPVSS-APA algorithm and convex combination APSA algorithm.
作者
火元莲
连培君
龙小强
王丹凤
HUO Yuan-lian;LIAN Pei-jun;LONG Xiao-qiang;WANG Dan-feng(College of Physics and Electronic Engineering,Northwest Normal University,Lanzhou 730070,Gansu,China)
出处
《西北师范大学学报（自然科学版）》
CAS
北大核心
2021年第6期57-62,93,共7页
Journal of Northwest Normal University(Natural Science)
基金
国家自然科学基金资助项目(61561044)。
关键词
自适应滤波
凸组合
最小均方算法
仿射投影算法
adaptive filter
convex combination
least mean square algorithm
affine projection algorithm
