Modeling and Model Predictive Control of a thermoacoustically unstable burner

Abstract

Abstract This paper describes the modeling and control of a thermoacoustic unstable burner. The thermoacoustic system is modeled using a 1D-acoustic network coupled with a flame transfer function derived from Computational Fluid Dynamics (CFD). The resulting unstable system provides a realistic test bench for controller evaluation and is controlled using a Model Predictive Controller (MPC) with a sampling time of 0.001s. As QP solver the homotopy approach of qpOASES is used.