Robust control based on neuro-fuzzy systems for a continuous stirred tank reactor

This paper studies the control problem of the concentration for a continuously stirred tank reactor (CSTR) in which parameter uncertainty and system disturbance are considered. A double control scheme, based on the PID control law and the internal model control strategy, is studied. Because the controller constructed by the neuro-fuzzy model is not very accurate and leads to control performance degradation, the double control scheme is proposed. The experiment study shows that the double control scheme adopted can extend effectively the controllable range and give good robust tracking control performances.