论文部分内容阅读
本文提出一种新颖的非线性自适应滤波器－递归神经网络（ＲＮＮ）自适应滤波器．讨论ＲＮＮ滤波器的构成及算法，实现了ＲＮＮ自适应有源噪声对消器．ＲＮＮ自适应有源噪声对消器性能优于内线性滤波器组成的线性对消器，能适应各种噪声环境．仿真结果充分证明了这一点．
This paper presents a novel nonlinear adaptive filter - recurrent neural network (RNN) adaptive filter. The structure and algorithm of RNN filter are discussed, and RNN adaptive active noise canceller is implemented. The performance of RNN adaptive active noise canceller is better than linear canceller consisting of linear filter, which can adapt to various noisy environments. Simulation results fully proved this point.