New partial update algorithms for multichannel active noise control with online secondary path modeling

For multichannel adaptive active control algorithms with online secondary path modeling, the computational load increases making real-time implementation impossible when the number of the reference sensors, secondary actuators and error sensors become large. In this paper, a new algorithm, Partial-update Filtered-e (PFeLMS) algorithm, is proposed to effectively reduce the computational load. In the proposed algorithm, the basic ideas for partial update weights and Filtered-e algorithm are combined and applied to online secondary path modeling and active control. The implementation approach for the proposed algorithm is given in detail. Finally using the PFeLMS algorithm, active control of magnetic resonance imaging (MRI) noise within a rectangular enclosure with time varying secondary paths is investigated by computer simulations, it is shown that considerable reduction in computational load as well as noise level can be achieved simultaneously.