Active noise control with dynamic recurrent neural networks.
Davor Pavisic, Laurent Blondel, Jean-Philippe Draye, G. Libert, Pierre Chapelle · 1995
. We have developed a neural active noise controller which performs better than existing techniques. We used a dynamic recurrent neural network to model the behaviour of an existing controller that uses a Least Mean Squares algorithm to minimize an error signal. The network has two types of adaptive parameters, the weights between the units and the time constants associated with each neuron. Measured results show a significant improvement of the neural controller when compared with the existing system. 1 Introduction. Active Noise Control (ANC) uses the intentional superposition of acoustic waves to create a destructive interference pattern such that a reduction of the unwanted noise occurs (Young's principle). Acoustic waves propagating in a rigid walled waveguide are one of the best candidates for ANC, first, because a considerable part of environmental noise is transmitted via ducts (ie. noise from ventilation systems or exhaust pipes), and second, because the sound wave in a duct...