Set-Point Supervisory Control Methodology for a Nonlinear Continuous Stirred Tank Reactor Process
Abstract
In this paper, a set-point supervisory control methodology for switching and tuning is developed for nonlinear systems. This control strategy is evaluated in the simulation of highly nonlinear continuous stirred tank reactor process. In the first step, closed-loop adaptive control algorithm is used to estimate a bank of models and to calculate their corresponding controllers, each of these models and their associate controllers are tuning around a given operating points. Thereafter, an intelligent supervisor is implemented to orchestrate the switching between different models and to select at each instant the adequate and the best controller to the plant. The simulation results found for three scenarios that correspond to different scaling of the set-point show the effectiveness and superiority in terms of performances and robustness characteristics of this proposed control approach comparing with closed loop adaptive controller.