A high level of process automation is a key to the safe, efficient, and competitive manufacture of high quality materials. A common focus of control research is the development of new technologies to build highly autonomous chemical processes; autonomous in the sense that they can function with a reduced level of operator intervention. The main issues to be dealt with in this research project are the integration of control and fault diagnosis to realize fault tolerant control systems, making effective use of process information, and developing a coherent theoretical approach to discrete process operations. The work will consist of three specific studies on new, fundamental approaches to process monitoring: (1) the integration of multivariable control and fault detection in L1/l using a four parameter controller; (2) a study of the application of multiresolution techniques for signal processing (wavelets) to process monitoring and control; and (3) the development of a novel hybrid continuous/discrete-event model for process fault detection and diagnosis. The study goals include the development of new theoretical approaches to these problems. Significant effort will be devoted to the development of effective computational procedures. Cooperation with active industrial research groups will foster the transfer of the technology to industrial practice.
