Intelligent process control with integrated robust multivariable controllers and real-time supervisory knowledge-based systems

The application of a supervisory knowledge-based system (KBS) to provide robust multivariable control of a chemical reaction process with reduced conservatism is examined. The supervisory KBS is capable of monitoring the process to detect process and control system faults or deteriorating control system performance due to changes in the process behavior or operating conditions. If a fault or change in performance is detected, the KBS formulates and implements the necessary corrective action. Control loop tuning is emphasized. The underlying mechanisms are discussed and validation results, using a rigorous simulation of the chemical reactor process, are presented.<
>