Multi-loop adaptive controllers in nonlinear process control

Adaptive controllers are investigated for control of a high-purity binary distillation column. A generalized predictive controller and an adaptive PID (proportional plus integral plus derivative) controller based on expert system concepts are used in a multiloop configuration to control column bottom and top compositions. The performance of these controllers is also compared with that of the conventional PID controller. The results obtained have shown that the adaptive PID controller based on expert system and pattern recognition ideas provides good results from the control, engineering, and cost-effectiveness points of view.<>