Design of a training based fuzzy controller for power plant de-superheaters

In this paper a new control method based on a combination of inverse dynamics method and neuro-fuzzy inference systems is developed for a nonlinear industrial plant. The method is applied to a super-heater system of a steam power generating plant. The controller's performance is compared with that of the existing PID feedback control system. A neuro-fuzzy model of this nonlinear plant is also developed based on the experimental data obtained from a complete set of field experiments. Comparing this nonlinear model with a linear model obtained from the least square error (LSE) method; it is shown that the neuro-fuzzy model is more accurate than linear model in the sense that its response is closer to the response of the actual system under different operating conditions. Comparison between the responses of the closed-loop control system under the proposed control strategy with the responses of the exiting control system shows the advantages of the new designed control system. It is demonstrated that with the proposed controller, the control system tracks the desired variable set points more accurately than the exiting PID controller.