摘要
基于变步长类最小均方(VS-LMS-L)算法和无记忆非线性梯度(MNG)算法,设计了一种变步长无记忆非线性梯度(VS-MNG)算法.新算法在频率估计均方误差(MSE)和收敛速度上获得增进的性能,而且具有很好的鲁棒性.仿真结果显示了设计算法的优越性.
A new gradient-based algorithm for adaptive Infinite Impulse Response (IIR) notch filters (VS-MNG) was proposed, which is based on a variable step-size LMS-like (VS-LMS-L) algorithm and memoryless nonlinear gradient (MNG) algorithm. The new algorithm has an improved performance in frequency estimation mean square error and convergence speed, and bears good robustness. Simulations are provided to show the superiority of the proposed algorithm.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2004年第12期1989-1991,共3页
Journal of Shanghai Jiaotong University
基金
国家自然科学基金资助项目(60172018)
关键词
无限冲击响应限波器
均方误差
梯度
自适应
infinite impulse response(IIR) notch filter
mean square error (MSE)
gradient
adaptive
