Extending a Combination of Map-based and Adaptive Feedforward Control Algorithms for Active Noise Cancellation

In order to reduce tonal orders of a combustion engine with electrical drives as acoustical actuators, a parallel control structure for active noise cancellation (ANC) applications is used. It consists of map-based and adaptive feedforward control algorithms to reduce the convergence time and improve the tracking behavior during fast changes of the disturbing motor order. This paper describes two extensions of the parallel control structure. First, a method to improve the noise reduction during the adaption process is proposed. Second, a method is described to constrain the filter weights of the parallel control structure to avoid nonlinear distortions due to output constraints.