基于约束反赫布算法的自适应二维FIR滤波器实现

The Realization of Adaptive 2-D FIR Filter Using Constrained Anti-Hebbian Algorithm

提出了一种用神经网络来实现自适应二维FIR滤波器的新方法,主要是运用约束反赫布学习算法,训练一个神经网络,实现了基于TLS准则的自适应二维FIR滤波器,计算机仿真结果表明该滤波器在收敛性能和精度上均优于基于LMS准则的自适应二维FIR滤波器。

This paper proposes a novel neural network algorithm for to realize an adaptive 2-D FIR filter, the main idea is to train a neural network with constrained anti-hebbian algorithm to realize the adaptive 2-D FIR filter based on TLS principle .Simulation resuits demonstrate that the filter has better performances of convergence than that of using LMS algorithm.