Fuzzy-PID Controller with Fuzzy Reasoned Adaptable Control Target and Its Application to Petroleum Plant in Transient State

In transient states of feed oil switching and operation mode changing, the nonlinear response with a time delay makes difficult to control a plant automatically. However, a well experienced operator can control it by his experience in such a state. We propose a hybrid control system by replacing the procedure of a well experienced operator. The proposed hybrid control system consists of three component; (1) the base control component which calculates an average from long-term tendency of the plant, (2) the compensating component which compensates the output of the base control component by predicting the status of the plant and (3) the target setting component which sets a suitable target value of controller. We choose PID controller for the base control component, fuzzy controller for the compensating component and fuzzy controller for the target setting component. This control system is applied to the tower-topp temperature control of the naphtha desulfurize plant at Idemitsu Hokkaidou Refinery and we have a good control performance in transient states of feed oil switching and operation mode changing.