Error Path Modeling of Dynamic Systems in Active Noise Control

Active noise control (ANC) systems based on adaptive algorithms typically require a prior setup stage to characterize the acoustic propagation between the control loudspeakers and the error microphones. Many algorithms are highly sensitive to errors in these estimates and exhibit limited robustness when the acoustic paths change over time, restricting their applicability to systems where the acoustic environment remains static or quasi-static during algorithm execution. In this paper, a novel ANC algorithm is proposed that requires no prior knowledge of the acoustic propagation paths between the system transducers. The required acoustic path information is obtained online during the control adaptation process without compromising convergence, and the proposed algorithm is demonstrated to be inherently robust against changes in the acoustic system.