New hybrid structure for adaptive active noise control

This paper proposes a new hybrid structure combining feedforward control and feedback control for active noise cancellation, and presents new adaptive algorithms for updating the coefficients of the two controllers. A stability-assured adaptive algorithm is investigated on the assumption that the secondary path dynamics is known, and further a fully adaptive algorithm is also given even when the secondary path dynamics is unknown as well. To reduce the actual canceling error, two virtual errors are introduced and are forced into zero by adjusting four adaptive FIR filters in an on-line manner, which can result in the canceling at the objective point. Unlike other conventional approaches, the algorithm does not need exact identification of the secondary path dynamics, and requires neither any dither signal nor the PE property of the source noise.