Multiple Channel Active Noise Internal Model Control with Saturation Nonlinearities

In active noise control (ANC) the nonlinearity sources could originate from the noise process, primary and secondary propagation paths, and actuators consisting of loudspeaker, microphone or amplifier. The saturation of the loudspeaker-amplifier in the secondary path is considered as the dominant nonlinearity. This paper proposes a nonlinear multiple channel filtered-x least mean square algorithm (NMFXLMS) for nonlinear active noise control (ANC) applications. A Hammerstein structure is used to model the secondary path. The linear part of the model represents the secondary acoustical propagation path while the nonlinear part represents the saturation nonlinearity of the loudspeaker amplifier. An internal model control (IMC) based feedback MIMO ANC system is used for the control structure. In comparison to the conventional linear multiple channel FXLMS algorithm, simulation results show that the proposed NMFXLMS algorithm improves the noise reduction performance in ANC systems with saturation nonlinearities.