Abstract - Kosanovich - 9703252 The objective this research is the development of a rigorous theoretical framework that can be used to design an effective feedback control strategy for multi-product chemical reactors. The motivation is provided by a global market economy which demands that the manufacture of multiple grades of a single product be made efficiently using the same equipment while simultaneously minimizing the production of off-spec material and maintaining safety and environmental constraints. This project is a fundamental approach to this problem that starts with modeling and analysis of the multiple nature of the operations leading to a framework that permits the design of a hybrid supervisory control strategy. Specifically: (1) local approximation models will be developed to represent the multiple process operations and constraints at each operating point, and (2) a dynamic supervisory controller will be designed to select and to switch into operation, the best controller from a set of possible controllers, that will provide superior setpoint tracking. This research will provide new and fundamental insights into the transition control problem, resulting in a new theoretical framework that combines multi-resolution modeling and robust controller design with analysis of stability and performance of the closed-loop system. The expected results will enhance the education of chemical engineers at the graduate and undergraduate levels. Examples of this include: the development of numerical programs in the form of algorithms to introduce multi-resolution model identification; the inclusion of transition control to the existing process control course; the development of a laboratory experiment to provide active experimentation and concrete experience of the proposed concepts; the collaboration with industry through co-operative externship programs to expose students to multi-product processes; and the use of workshops and thesis projects with industrial students to facilitate technology transfer.
