Enhancement of active noise control using neural-based filtered-X algorithm

This paper presents a neural-based filtered-X least-mean-square algorithm (NFXLMS) to cancel the nonlinear broadband noise in an active noise control (ANC) system. The ways to avoid the premature saturation of backpropagation algorithm and to design the optimal learning rate are also included in the paper to improve the noise reduction performance. Besides, the proposed neural filter can be easily implemented and versatile to the other applications. Several simulation results show that the proposed method can effectively cancel the narrowband and nonlinear broadband noise in an ANC system.