In industrial processes and many other control problems, even now more than 95% of the controllers are of the PID controller type. The PID controller has several important functions. It has ability to eliminate steady-state error through integral action and anticipate the future through derivative action. The PID controller, therefore, can not only track the given set-point but reject the disturbance that enters the control loop. This is the most basic reason that there exist many different methods for tuning the PID controller.
In this thesis, tuning rules of the PID controller for load disturbance rejection are proposed incorporating with real-coded genetic algorithms(RCGAs). The optimal parameters sets of the PID controller are obtained based on a first-order plus time delay model and a RCGA. As for assessing the performance of the controller, three performance criteria (ISE, IAE and ITAE) are adopted. Then tuning formulae are derived using the tuned parameters sets, potential tuning rule models and another RCGA.
A set of simulation works are carried out on three high order processes with time delay to verify the effectiveness of the proposed rules.