Bayesian active noise control

Abstract

Active Noise Control (ANC) is a challenging practical application of adaptive control systems. This paper approaches ANC from the perspective of the Bayesian Inverse Problems theory. The ANC underlying problem is initially formulated as a generic Bayesian inverse problem. A solution to this problem is then obtained using standard methods in the Bayesian Inverse Problems theory, resulting in a new adaptive algorithm for ANC. The results show the effectiveness of the chosen approach in creating new adaptive algorithms for ANC, one of which is presented in this paper. This algorithm can reach a probabilistic model for the optimal control systems, but conventional algorithms can reach only a deterministic model. Consequently, unlike other algorithms, the proposed algorithm can quantify the uncertainty associated with the adaptive control process in active noise control.