用于FIR滤波器的递归最小二乘拟牛顿算法A New Recursive Least Squares Quasi-Newton Algorithm for Finite Impulse Response Filters
王卓,冯大政
摘要(Abstract)：
基于拟牛顿优化方法,提出了一种稳健的自适应FIR滤波算法。新算法用最小二乘误差(LSE)代替了均方误差(MSE)作为代价函数,他具有和常规递归最小二乘(CRLS)算法相近似的追踪能力,且不存在数值计算不稳定性的问题,在收敛速度以及稳态效果方面也要优于DeCampos的拟牛顿(QN)算法。由计算机仿真比较了有关算法的性能。
关键词(KeyWords)： 自适应滤波;拟牛顿;FIR滤波;递归最小二乘
基金项目(Foundation): 国家自然科学基金资助项目(60372049)
作者(Author): 王卓,冯大政
参考文献(References)：
- [1]Haykin S.Adaptive Filter Theory[M].4th Edition.Engle wood Cliffs,NJ:Prentice Hall,2001.
- [2]De Campos ML R,Antoniou A.A New Quasi Newton Adaptive Filtering Algorithm[J].IEEE Transactions on Circuits and Systems II:Analog and Digital Signal Process ing,1997,44(11):924934.
- [3]Da Zheng Feng,Xian Da Zhang,Dong Xia Chang.A Fast Recursive Total Least Squares Algorithmfor Adaptive FIR Filtering[A].IEEE Transactions on Signal Processing,2004,52(10):27292737.
- [4]Verhaegen M H.Round off Error Propagationin Four Generally Applicable,Recursive,Least squares Esti mation Schemes[J].Automatica,1989,25(3):437444.
- [5]Bottomley G E,Alexander S T.A Novel Approach for Sta bilizing Recursive Least Squares Filters[J].IEEE Transac tions on Signal Processing,1991,39(8):17701779.
- [6]Yuexian Zou,Shing Chow Chan.A Robust Quasi Newton Adaptive Filtering Algorithmfor I mpulse Noise Suppression[A].IEEEISCAS2001,Sydney,Australia.2001,2:677680.
- [7]胡塘,刘文波,于盛林.FIR滤波器的一种新型设计方法[J].现代电子技术,2004,27(21):3233,41.
- [8]陈开周.最优化计算方法[D].西安:西北电讯工程学院,1984.