Fuzzy Supervisory Sliding-Mode and ANFIS Control for Drum water level of Boiler

One of the main parameters in steam power plant is water level control of drum to prevent drum components over- heating or steam lines flooding. The proportional integral (PI) controller approach is a conventional method for drum level control. Due to disturbances, model uncertainties and parameter mismatch this control scheme (PI) not guaranty the proper and robust performance of the system. A nonlinear control system must be developed to achieve the appropriate performance. Sliding mode controller is a robust state feedback controller that has fast transient response but because of chattering it is not efficient in steady state condition. On the other hand, the Adaptive neuro fuzzy inference system (ANFIS) has been applied for drum level control. However, through inevitable training process this controller has not proper behavior in dealing with uncertainties. In order to use advantages of both controllers in this paper combination of both sliding mode and Adaptive neuro- fuzzy inference system with variable weight have been used. Also a fuzzy supervisor is designed to determine the weight of each controller. In addition, fuzzy supervisor causes smooth switching between two controllers.

Keywords:

Steam drum water level control , Sliding mode control , ANFIS , Fuzzy supervisor