Active control of vehicle powertrain noise using adaptive notch filter with inverse model LMS algorithm

Conventional active noise control (ANC) systems are typically configured with the filtered-x least mean squares (FXLMS) algorithm or its modified versions. However, the traditional FXLMS algorithm often exhibits a frequency-dependent convergence behaviour, which leads to a poor tracking ability and unbalanced performance at individual harmonics. In this study, a novel adaptive notch filter with inverse model least means squares (ANF-IMLMS) algorithm is proposed as the basis for active control of vehicle powertrain noise. The proposed algorithm possesses the following two salient features as compared to the filtered-x LMS type algorithms: 1) rapid convergence speed; 2) good computational efficiency. The convergence speed and computational complexity of the proposed algorithm is analysed first. Then, the proposed ANC system for vehicle powertrain noise configured with the new algorithm is evaluated. The results show obvious enhancement in the convergence speed and noticeable noise reductions for each engine harmonic over a broader frequency range.