An active noise controller using type-2 fuzzy neural network

Abstract

This paper presents a feedback active noise control system using type-2 fuzzy  neural network. The  new features of  the proposed system are: firstly we introduce  a saturation compensator for the actuator and secondly we use a type 2 fuzzy neural network to estimate the nonlinearity of the secondary path transfer function. Online dynamic back-propagation learning algorithm based on the error gradient descent method is proposed. The condition of convergence of  the proposed  algorithm is derived using a discrete Lyapunov function. Simulation results show that the proposed method is effective.