A narrowband active noise control system with autoregressive model and linear cascaded adaptive notch filter

• The frequency estimation algorithm based on linear cascaded adaptive notch filters (LCANF) combined with autoregressive (AR) model is proposed. • The propose algorithm can estimate frequencies with the provided initial value of 0 Hz. • A time-sharing coordination mechanism for the LCANF, the AR model and the ANC subsystem is designed in the new NANC system. • With the reference frequencies accurately refined by the AR model, the new NANC system has with low computational load and fast convergence, and reaches an excellent steady-state noise reduction performance. Active noise control (ANC) system outperforms traditional passive noise control system in suppressing narrowband noise with low frequency. To reach a superior control quality, an accurate frequency of reference signal is required for the narrowband ANC (NANC) system. The existence of frequency mismatch (FM) degrades the performance of the system. In this paper, a frequency estimation algorithm based on linear cascaded adaptive notch filter (LCANF) combined with the autoregressive (AR) model is designed for the NANC system, which is simple and feasible as well as has an extremely low computational load. In the new NANC system, the rough estimations of multiple tones are given by the LCANF algorithm. Then the results are fed into the AR models, which fine-tune the estimated frequencies and offers the reference signals to the NANC system. The designed new mechanism effectively cuts down the computational load. Extensive simulations were conducted and the results demonstrate that the proposed system converges fast and has prominent noise reducing performance when it deals with periodic noises.