Supervisory control of Discrete Event Systems is a relatively new research area motivated by the pragmatic need to solve the problem of modeling and control of large-scale complex systems in discrete-event setting. The pertinent literature on Discrete Event Supervisory (DES) control offers the innovative idea of language measure to quantitatively evaluate the performance of a DES controller. This dissertation addresses fundamental issues about language measure and optimal DES control to initiate a benchmark control strategy which is not available in existing literature. The objective of this dissertation is to enhance the theory and application aspects of the DES control in a hybrid perspective. The new optimal control strategies along with the enhanced event generation algorithms are developed and used as benchmark to fully elucidate and improve the optimal control and event generation concepts in the literature of hybrid supervisory control.

The integrated experimental and analytical studies presented in this dissertation envelop the implementation of the new theories of DES control in a hierarchical and hybrid fashion that will incorporate continuously-varying and discrete-event dynamics. Emerging applications of hybrid supervisory control to complex dynamical systems such as gas turbine engines, rotorcrafts and mobile robotic platforms are investigated, and validation of the theoretical work by simulation and experimentation is performed. Continuous and DES controllers are implemented in a hybrid setting to regulate both system dynamics and operational behaviors. Optimal DES controllers are designed for a twin-engine propulsion system which is integrated with an aircraft model and flight control, and a Segway robot which autonomously performs coordinated missions together with an unmanned aerial vehicle.

It is envisioned that this dissertation will make a positive contribution toward successful implementation of the hierarchically structured DES decision and control theory to many complex systems.