New feedback active noise control system with improved performance

In many practical active noise control (ANC) applications, the undesired primary noise consists of multiple harmonic-related tones, which can be effectively reduced by internal model control (IMC) feedback ANC systems. In this paper, a new feedback ANC system is proposed with improved convergence rate and noise reduction when the frequency separation between two adjacent tones is small. In the proposed system, adaptive notch filter (ANF) is used to estimate the frequencies of the multi-tone noise, based on which the reference signals are internally generated and regrouped to increase the frequency separation in each channel. Compared with the conventional IMC feedback ANC system, the proposed system converges faster, has a better tracking capability when the frequencies of the primary noise vary, and is also less sensitive to impulsive noise. Computer simulations are conducted for both synthesized tonal noise and motorcycle engine noise to show the improved performance of the proposed system.