Design of a Data-Oriented Performance Driven Control System Based on the Generalized Minimum Variance Control Law

In process industries, it is necessary to maintain the user-specified control performance in order to achieve desired productivity. This paper describes a design scheme for control performance evaluation based PID controllers in which the "controller design" is driven by a performance assessment" scheme. According to the proposed scheme, controller parameters are tuned using a fictitious reference iterative tuning (FRIT) scheme. In FRIT, controller parameters are calculated by using only closed-loop data. The proposed scheme is a type of data-oriented adaptive controller whose controller parameters are tuned directly without requiring a model of the process. Unlike previous performance-based adaptive schemes, the control objective function to be minimized takes into account controller error variance as well as the manipulating variable or the control effort variance. The effectiveness of the proposed scheme is verified by using a simulation example and the benchmark stirred tank-heater system.