基于遗传算法的自适应滤波器设计与实例仿真  

Self-adaptive filter design and its simulated application example based on genetic algorithm

作　　者：王晓芳[1] 

机构地区：[1]山东轻工业学院电子信息与控制工程学院,山东济南250100

出　　处：《山东轻工业学院学报（自然科学版）》2006年第2期24-28,共5页Journal of Shandong Polytechnic University

摘　　要：LMS算法是自适应滤波器最为常用的算法,该算法比较简单,但决定其收敛速度和稳定性的学习速率难以确定;遗传算法是一种高度并行的全局搜索方法,能够有效地用于自适应滤波器的权系数寻优。文章讨论了基于遗传算法的自适应滤波器的设计过程,并作为一个应用实例,将其应用于线性系统辨识中,取得了较好的仿真结果。The Least Mean Square Algorithm is the most common algorithm used in self-adaptive filtering. That algorithm is simple, but the learning rate that decides the speed and stability of learning process, is difficult to determine. Genetic Algorithm is a highly parallel and global searching method, it can also be used to search the optimum of the self-adaptive filter weights. In this paper the self-adaptive filter design process based on genetic algorithm is discussed, then applied to linear system identification as an example and satisfactory simulated results were obtained.

关 键 词：遗传算法 自适应滤波 系统辨识 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]