| 模糊步长LMS算法及其性能分析 |
| |
| 引用本文： | 林川,冯全源.模糊步长LMS算法及其性能分析[J].系统工程与电子技术,2007,29(6):967-970. |
| |
| 作者姓名： | 林川 冯全源 |
| |
| 作者单位： | 西南交通大学信息科学与技术学院,四川,成都,610031 |
| |
| 摘 要： | 以均方误差、输出与误差信号的相关系数作为衡量LMS算法收敛程度的标准及模糊推理系统的输入,提出了一种用零阶Sugeno模糊推理系统自适应调整步长的模糊步长LMS(FSS-LMS)算法,并从理论上分析了FSS-LMS算法的计算复杂度及其收敛性能。分析结果指出FSS-LMS算法的计算复杂度与传统LMS算法基本相当,但它具有更大的灵活性。自适应系统辨识的仿真结果表明FSS-LMS比传统的LMS算法及其它一些变步长LMS算法具有更好的收敛性能。
|
| 关 键 词： | 模糊推理系统 模糊步长 LMS算法 相关系数 自适应滤波 |
| 文章编号： | 1001-506X(2007)06-0967-04 |
| 修稿时间： | 2006年5月18日 |
Fuzzy step size LMS algorithm and its performance analysis |
| |
| LIN Chuan,FENG Quan-yuan.Fuzzy step size LMS algorithm and its performance analysis[J].System Engineering and Electronics,2007,29(6):967-970. |
| |
| Authors: | LIN Chuan FENG Quan-yuan |
| |
| Abstract: | With the mean square error,the correlation coefficient between the output and error signal as the input of fuzzy inference system and a measure of convergence,a fuzzy step size LMS(FSS-LMS) algorithm is proposed.The complexity and convergence performance of FSS-LMS algorithm are also analyzed.FSS-LMS algorithm uses zero-order Sugeno fuzzy inference system to adjust the step size of LMS algorithm.Its complexity is comparative to the traditional LMS algorithm but has better flexibility.Simulation results show that FSS-LMS has superior convergence performance than conventional LMS and some other variable step size LMS algorithms. |
| |
| Keywords: | Fuzzy inference system Fuzzy step size LMS algorithm Correlation coefficient Adaptive filtering |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
