基于RBF神经网络的非线性滤波器的研究
首发时间:2011-04-21
摘要:数字滤波器设计是数字信号处理领域一个重要研究方向。而当前的数字滤波器存在不能对滤波器的权系数进行实时计算、不能实时实现非线性滤波等问题。针对上述问题,本文提出了一种非线性滤波方法,即利用RBF网络所唯一具有的精确地局部逼近特性,将其应用到非线性滤波。通过MATLAB软件进行网络的设计与误差分析可知,将RBF网络运用到非线性滤波是可行的而且效果较好。该方法弥补了线性滤波器的对非线性干扰的处理缺陷,改进了以往的滤波技术。
For information in English, please click here
Research on Nonlinear Filter Based on the RBF Neural Network
Abstract:Digital filter design plays an important role in the digital signal processor field. But current filtering exist many problems such as can not calculate weighting coefficients of filter in real time and achieve real-time nonlinear filtering. In order to solve the problems proposed above. A nonlinear filtering method which utilizes local approximation properties of RBF is proposed in this paper,And has applied it to nonlinear filtering. Through MATLAB network design and error analysis we can see that applied RBF to nonlinear filtering is a feasible and get a better result. The method can make up the processing defect of linear filter when dealing with the nonlinear interference and improve previous filtering technology.
Keywords: RBF Non-linear Filter signal process
基金:
论文图表:
引用
No.4422151583620130****
同行评议
共计0人参与
勘误表
基于RBF神经网络的非线性滤波器的研究
评论
全部评论0/1000