基于BA优化核参数的非线性Volterra滤波方法研究  

Research on Nonlinear Volterra Filter Method with Kernel Parameters Optimized Based on BA

作　　者：赵玮 王强 何晓晖 袁媛 ZHAO Wei;WANG Qiang;HE Xiao-hui;YUAN Yuan(National Defense Engineering College,Army Engineering University of PLA,Nanjing 210007,China;Field Engineering College,Army Engineering University of PLA,Nanjing 210007,China)

机构地区：[1]陆军工程大学国防工程学院,江苏南京210007 [2]陆军工程大学野战工程学院,江苏南京210007

出　　处：《仪表技术与传感器》2019年第6期77-81,共5页Instrument Technique and Sensor

基　　金：国家自然科学基金面上项目(51175511)

摘　　要：针对传统的滤波方法对混合噪声滤除性能不足的问题,提出一种针对混合噪声的基于Volterra 级数的非线性滤波方法,采用蝙蝠算法对Volterra滤波器的核参数进行优化。仿真实验结果表明,经蝙蝠算法优化核参数的非线性Volterra滤波器能够较传统滤波方法有效滤除混合噪声,信号失真现象不明显,同时其运算量小的特点较传统滤波方法具有更好的实时性。Aiming at the problems of the bad filter performance for mixed noise in traditional filter method,a nonlinear filte ring method based on Volterra series was proposed,the kernel parameters of Volterra filter were optimized by bat algorithm.The simulation results show that nonlinear Volterra filtering with kernel parameters optimized by bat algorithm can more effectively fil ter mixed noise than traditional filter,the distortion of output signal was not obvious,and the new filtering method with less calcu lated amount was more real time than traditional filter method.

关 键 词：信号降噪 混合噪声 Volterra滤波 核参数 函数优化 蝙蝠算法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]