Active Noise Control Using a Feedforward Network with Online Sequential Extreme Learning Machine

Abstract

In practical active noise control (ANC) systems, the primary path and the secondary path may be nonlinear. The actuators of an ANC system often have non-minimum phase response. A linear controller under such situations yields poor performance. Neural networks using Filtered-x back-propagation (FX-BP) algorithm are often used as a controller for the nonlinear ANC systems. But FX-BP algorithm often converges slowly and may converge to a local minimum. A novel feedforward network-based ANC algorithm is proposed in this paper. The Online Sequential Extreme Learning Machine(OS-ELM) is generalized to meet the requirements of the nonlinear ANC systems. Computer simulations have been carried out to demonstrate that the proposed algorithm outperforms the FX-BP algorithm when the primary path is nonlinear.