Adaptive Hammerstein filters for nonlinear and non-minimum phase active noise control systems

In this paper, the adaptive Hammerstein Filtered-X LMS (HFXLWS) filter is introduced and applied to ANC systems. The proposed algorithm is compared to the conventional Filtered-X LMS (FXLMS) algorithm as well as the Volterra Filtered-X LMS (VFXLMS) algorithm in two different experiments; a) a nonminimum phase secondary-path transfer function, and b) a nonlinear primary-path transfer function. Using the simulation results, it is shown that the developed algorithm achieves a lower mean-square error (MSE) with respect to the KYLAE algorithm, and almost a comparable convergence rate and MSE compared to the VFXLMS algorithm, but with a simpler structure.