9414780 Garg As man-made systems such as automated factories, communication networks, and computers continue to grow in size, "cut and try" approaches to design and supervision become less realistic. What is needed instead are controller synthesis techniques that allow optimal control laws to be derived mathematically and with performance properties that can be verified automatically. A well-established framework exists for control of systems without explicit real-time constraints. However, if timing is of concern in the system dynamics and its performance specification then the situation becomes much more complex. Certain real-time discrete event systems can be described by linear equations in a non-traditional algebraic system called max-algebra. For these systems, the problem of supervisory control is to impose delays on the execution of selected events to achieve some specified performance goal. The objective of the proposed research is to examine the existence and synthesis of controllers for DES which include such time constraints. Our method relies on the max-algebra model for timed DES which permits control theory results for untimed DES to be related to timed systems. ***
