Thermoacoustic instability: Model-based optimal control designs and experimental validation

Active control of thermoacoustic instability has been increasingly sought after in the past two decades to suppress pressure oscillations while maintaining other performance objectives such as low NOx emission, high efficiency, and power density. Recently, we have developed a feedback model of a premixed laminar combustor which captures several dominant features in the combustion process such as heat release dynamics, multiple acoustic modes, and actuator effects. In this paper, we study the performance of optimal control designs including LQG-LTR and H-infinity methods using the model with additional effects of mean heat and mean flow, actuator dynamics, and input saturation. We also verify these designs experimentally using a 1 kW bench-top combustor rig and a 0.2-W loudspeaker over a range of flow rates and equivalence ratios. Our results show that the proposed controllers, which are designed using a two-mode finite dimensional model, suppress the thermoacoustic instability significantly faster than those obtained using empirical approaches in similar experimental setups without creating secondary resonances, and guarantee stability robustness.