Neuro-fuzzy tuning of PID controller for control of actual gas turbine power

Abstract

The purpose of introducing combined cycle with gas turbine in power plants is to reduce losses of energy. Their main role lies in the utilization of waste heat that may be found in exhaust gases from the gas turbine or at some other points of the process to produce additional electricity. The efficiency of the plant increases reaching over 50%, while the traditional steam turbine plants is approximately 35%/spl ap/40% or so. Up to date, the PID controller has been used to operate under such systems, but since the gain of PID controller manually has to be tuned by trial and error procedures. Getting an optimal PID gains is very difficult to tune manually without control design experience. In this paper, we studied an acquiring of transfer function from operating data of Gun-san gas turbine in Korea and a new 2-DOF PID controller tuning by NFS is designed for the optimum control to Guns-san gas turbine's variables variety. Since the shape of a membership function in the NFS vary on the characteristics of plant. ANFIS based control method is effective for plant that their variables vary. Its results are compared to the conventional PID, 2-DOF PID controller and represents satisfactory response. We expect this method will be used for another process because it is studied on the real operating data.