Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Electrical and Computer Engineering

First Advisor

Kemin Zhou


This dissertation presents some advances in active control of thermoacoustic instabilities in combustion chambers. Large-size gaseous and liquid fueled swirl stabilized combustors were used during the studies. Active control was implemented using different types of actuators. Proportional (loudspeakers and fuel valves) and discrete actuators (open-close automotive fuel injectors) were investigated. Acoustic and fuel modulation control were successfully applied. In large-scale combustors, flame stabilization techniques such as swirl add three dimensional characteristics to the flow. Moreover, the induced turbulence creates highly nonlinear interactions in the system. Thus, in order to capture these characteristics nonlinear partial differential equations have to be used. Alternatively, the main dynamics of the combustion process can be modeled experimentally. This approach was chosen. Time and frequency domain linear identification techniques were used for this purpose. Several model-based control strategies such as LQG, Hinfinity Disturbance Rejection and Hinfinity Loop-Shaping techniques were tested experimentally with success. A simple controller whose parameters were optimized on-line is also introduced. An evolution algorithm was developed to perform its parameter optimization achieving good convergence to optimal values. The improvements with these proposed control techniques over classical phase-delay control are demonstrated experimentally. A new control configuration was suggested from heat-release visualizations of the flame. In this new configuration, control actuation is directly focused onto the main area of heat-release in the flame front. Consequently, a more efficient actuation is achieved. It is shown that with just a small amount of modulated fuel, phase-delay control can substantially attenuate the pressure oscillations. Finally, during the development of Hinfinity controllers, there were cases where the stability of the resulting controllers restricted the closed-loop performance. A control design strategy to solve the Hinfinity Strong Stabilization problem is then presented. The proposed design strategy pursues to overcome the conservativeness of existing formulations. Examples show its potential for future applications.