A Novel ANC System using Nonlinear Error LMS Algorithm

The conventional Filtered-x Least Mean Square (FxLMS) shows poor performance in the presence of non linearities present in the primary and the secondary path. The functional link artificial neural network (FLANN) has been successfully applied to tackle the nonlinearity in the primary and secondary paths. However, all these methods of active noise control use a linear model to obtain an estimate of the non linear secondary path. This paper proposes a FLANN based nonlinear model to obtain an estimate of thenon linear secondary path. In addition, a FLANN based controller is also developed to generate anti noise. A new update rule nonlinear-error least mean square (NELMS) is derived and used for updating the weights of the first FLANN model by utilizing the intermediate signals of the second FLANN model. The simulation results presented exhibit superior performance of the proposed method compared to that of FsLMS-ANC method.