Performance analysis of nonlinear active noise control system using weighted FLANN algorithm

Abstract

Functional Link Artificial Neural Network is popularly used as controller in nonlinear Active noise control system as it is characterized by simple structure and relatively low computational complexity. Learning capability of the controller is achieved by filter-s least mean square algorithm and its variants. But these class of algorithms are characterized by slow convergence. This paper uses a set of combining weights to expedite the convergence process and the performance of the active noise controller is analyzed by tracking the combining weights. Computer simulation experiments are also conducted and analysis results presented.