A COMPARISON OF THE GENERALIZED FOURIER SERIES AND FUNCTIONAL LINKS ARTIFICIAL NEURAL NETWORKS FOR NONLINEAR ACTIVE NOISE CONTROL SYSTEM

Abstract

The behavior of nonlinearity in the active noise control (ANC) system is not the same. Therefore, in order to increase the efficiency of noise reduction, we need to understand the type of nonlinearity in the ANC system and choose the appropriate model. This paper presents a comparison and evaluation between the even mirror Fourier series (EMF) and the functional links artificial neural networks (FLANN) for the nonlinear ANC system. By analyzing the nonlinear influences that exist in the primary path, the secondary path, and the noise source in the active noise control system, various types of nonlinearity, such as memory nonlinearity, memory-less nonlinearity, and chaotic nonlinearity has been discussed. Furthermore, the modeling capabilities of the expansion functions based on the EMF and FLANN for the types of nonlinearities have been analyzed. The causes for such behavior have also been pointed out. Many computational simulations in different nonlinear scenarios have been carried out to demonstrate the analysis and evaluation of ANC systems based on the EMF and FLANN models.