自适应加权FIR-Myriad混合滤波算法

Adaptive weighted FIR-Myriad hybrid filter algorithm

针对FIR滤波器滤除脉冲噪声以及加权Myriad滤波器滤除高斯噪声的不足,提出基于FIR滤波器和加权WMy滤波器有效组合的一类新的非线性滤波器FIR-WMyH滤波器。利用神经网络中的反向传播算法,在均方误差准则下,推导了一个基于统计梯度的自适应算法。基于稳定α分布脉冲噪声模型下的仿真结果说明了该算法的良好的性能。

A new class of nonlinear filters called as the weighted FIR-Myriad hybrid （FIR-WMyH） filters were proposed. These filters can overcome the disadvantages of FIR filter and weighted Myriad filter. Through the back-propagation algorithm used in neural networks, a stochastic gradient-based adaptive algorithm for determining optimal FIR-WMyH filters under the mean square error （MSE） criterion is derived. In the case of the stable distributed impulsive noise model, the good performance of this adaptive algorithm is demonstrated through the computer simulation results.