Design of a smart adaptive control system based on control performance evaluation

PID controllers have been widely used in a many process systems because of simple control structure and clear physical meaning of PID parameters. Among the PID tuning schemes, the VRFT (Virtual Reference Feedback Tuning) and the FRIT (Fictitious Reference Iterative Tuning) which determine control parameters by using closed-loop data have received much attention in recent years. This paper studies how to accomplish “controller design” and “control performance assessment” concurrently based on the GMVC (Generalized Minimum Variance Control). According to this scheme, PID parameters are directly determined by using a set of operating data. The effectiveness of the proposed method is evaluated by a simulation example.