本文在Ｈｏｐｆｉｅｌｄ神经网络和ＴＨ线性优化网络的基础上，提出了自适应滤波算法（ＬＭＳ和ＬＳ算法）的神经网络实现，理论和计算机模拟均表明该算法的神经网络硬件实现，将大大提高自适应滤波器的时效性． Based on the Hopfield neural network and TH linear optimization network, this paper proposes the implementation of neural network based on adaptive filtering algorithm (LMS and LS algorithm). The theory and computer simulation show that the neural network hardware implementation of this algorithm will greatly enhance the adaptive Timeliness of the filter.