Comparison of Feed-Forward and Recurrent Neural Networks in Active Cancellation of Sound Noise

Passive techniques such as barriers, silencers and isolation are bulky, costly and ineffective at low frequencies. Active cancellation of noise was presented because of these problems. In this paper, we want to investigate the uses of neural networks in active noise control (ANC). Feed-forward and recurrent neural networks are compared for active cancellation of sound noise. In order to compare the two networks the number of layers and neurons are equal in both of the networks. Moreover, training and test samples are similar for networks. The noise signals that are used for training the networks are selected from SPIB database. The results of simulation show the ability of these networks in noise cancellation. As it is seen, recurrent neural network has better performance in noise attenuation than the feed-forward neural network.