Active Noise Control for Harmonic and Broadband Disturbances Using RLS-Based Model Predictive Control

Abstract

This paper develops RLS-based MPC (RLSMPC), which uses multiple implementations of recursive least squares (RLS) to perform model predictive control (MPC). RLSMPC uses output-feedback measurements rather than full-state-feedback to construct the control input, thus removing the need for state estimation. To remove the need for an a priori model, RLSMPC uses RLS to perform online, closed-loop identification. This approach is applied to active noise control with unknown sinusoidal and broadband disturbances.