基于LMS自适应滤波器论文.doc

吉林化工学院毕业设计说明书 基于RLS的自适应滤波器的设计与实现 Design and Implementation of Adaptive Filter Using RLS 学生学号： 学生姓名： 胡金龙 专业班级： 电信0701 指导教师： 杨 明 职 称： 助 教 起止日期：2011.3.22～2011.6.11 吉 林 化 工 学 院 Jilin Institute of Chemical Technology 摘 要 自适应滤波器是统计信号处理的一个重要组成部分。在实际应用中，由于没有充足的信息来设计固定系数的数字滤波器，或者设计规则会在滤波器正常运行时改变，因此我们需要研究自适应滤波器。凡是需要处理未知统计环境下运算结果所产生的信号或需要处理非平稳信号时，自适应滤波器可以提供一种吸引人的解决方法，而且其性能通常远优于用常方法设计的固定滤波器。此外，自适应滤波器还能提供非自适应方法所不可能提供的新的信号处理能力。 本文从自适应滤波器研究的意义入手，介绍了自适应滤波器的基本理论思想，具体阐述了自适应滤波器的基本原理、算法及设计方法。自适应滤波器的算法是整个系统的核心。本文中，对两种最基本的自适应算法，即递归最小二乘（RLS)和最小均方误差(LMS)算法进行了详细的介绍和分析，并针对两种算法的优缺点进行了详细的比较。同时，分别对FIR结构和IIR结构自适应滤波器做了详细的介绍，比较了FIR结构和IIR结构自适应滤波器的优缺点。最终采用改进的RLS算法设计FIR结构自适应滤波器，并采用MATLAB进行仿真。实验结果表明，该自适应滤波器滤波效果优越。 关键词：自适应滤波器算法FIR结构滤波器Abstract The auto-adapted filter is an important part of the digital signal processing. In practical application, there haven’t sufficient information to design the fixed coefficient digital filter, or the design rule will be changed when the filter normally operated. Therefore, we need to research the auto-adapted filter. Whenever needs to process the signal under the unknown statistical environment, or the non-steady signal, the auto-adapted filter can provide a appealing solution. Moreover, its performance usually over the normally fixed filter. In addition, the auto-adapted filter can also provide the ability of the recent signal process which the non-auto-adapted method is impossible to provide. Firstly this thesis proposed the importance of auto-adapted filter research and introduced its elementary theory , algorithm and design method. The core of the whole system is the auto-adapted filters algorithm . In this article, two of the most basic auto-adapted algorithms, the smallest mean error (LMS) algorithm and the recursive least squares (RLS) algorithm, have particularly introduced and analyzed. Meanwhile, the advantages and disadvantages of this two algorithms have been compared. At the