基于自适应算法的去噪滤波仿真比较  

Comparison of Noise Filtering Simulations Based on Adaptive Algorithms

作　　者：赵薇[1] 陆余恬[1] 

机构地区：[1]中国传媒大学信息工程学院,北京100024

出　　处：《中国传媒大学学报（自然科学版）》2013年第4期40-46,共7页Journal of Communication University of China：Science and Technology

摘　　要：介绍了自适应滤波器的基本原理,对最小均方算法(LMS)和递推最小二乘算法(RLS)这两种自适应算法进行了理论分析和仿真比较。模拟仿真结果和语音去噪系统均表明:两种算法都能有效滤除噪声干扰。相比之下,RLS算法具有更好的收敛性能和稳定性,权噪声小,抑噪能力强。article introduces the basic principle of adaptive filter and the least mean squares algorithm （LMS） and recursive least squares （RLS） algorithm were applied to the theoretical analysis and simula- tion comparison. Both simulation results and speech denoising system indicate that the two kinds of algo- rithm can effectively filter out noise interference. In contrast, the convergence performance is better and the behavior of the RLS algorithm is much more stable, and it has lower misadjustment noise and better robustness aganist noise and disturbance.

关 键 词：自适应滤波 LMS算法 RLS算法 仿真 

分 类 号：TP273.21[自动化与计算机技术—检测技术与自动化装置]