Design of fuzzy control system for chemical injection systemretrofit using neural network model in thermal power plant

摘要

Many conventional thermal power plants were started up and shutndown infrequently and the operating loads were constant at the normalnoperation. In contrast, in recent years, with the participation ofnnuclear power generation and from the varied pattern of loadnconsumption, it has been required that the plants should be operatednefficiently. Thus, the DSS (daily startup shutdown) and WSS (weeklynstartup shutdown) have been carried out, and the plant startup, shutdownnand load variations have become very frequent. This tendency isnreflected also on water conditioning systems. The Poryoung thermal powernplant designed with the concept of DSS operation, and also, the CISn(chemical injection system) as one of water conditioning systems havenautomatic operation mode to maintain constant water quality withnconventional control system. Unfortunately, the control system does notnmaintain the set point of water quality. The engineer revise the controlnconcept with F(x) function, but the results of test operation do notnshow satisfactory performance. This paper discussed the parameternidentification of CIS using neural network model and suggested a fuzzyncontrol system for CIS to improve the control performance and showed thenvalidation of suggested system using simulation