一种基于支持向量机建模的新型自适应滤波器  

A new Widrow adaptive filter based on supportive vector machine

作　　者：刘斌[1] 李卓[2] 刘铁男[1] 喻胜阳 任珍珍[4] 

机构地区：[1]大庆石油学院电气信息工程学院,黑龙江大庆163318 [2]大庆石油学院地球科学学院,黑龙江大庆163318 [3]吉林石油集团公司人事部,吉林松原138000 [4]大庆石油学院外语系,黑龙江大庆163318

出　　处：《大庆石油学院学报》2005年第4期64-66,共3页Journal of Daqing Petroleum Institute

摘　　要：在Widrow B等人的自适应逆控制理论的基础上,提出了一种基于支持向量机的自适应滤波器.介绍了支持向量机的基本理论,结合支持向量机算法给出了该滤波器的原理图,讨论了滤波器的工作过程.计算机仿真实例表明,时变噪声的污染已被完全消除,新型自适应滤波器效果明显.This paper proposes a new adaptive filter based on support vector machine The new adaptive filter is based on the adaptive inverse control which is put forward by Widrow. Firstly, the basic principle of SVM is introduced. And the paper analyses the structure of the new adaptive filter based on SVM. Then the processing of the filter is discussed in detail. Lastly,, the time-varied noise has been filtered. The experiment given has proved the validity of the new algorithm.

关 键 词：统计学习理论 支持向量机 自适应滤波器 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]