Application of fuzzy-rough control in deoxidation system
Yanbin Qu, Jianyong Su, Li-Guo Feng · 2005
A multidimensional fuzzy controller is presented based on rule reduction of rough sets. The design of the controller is simplified by reducing the multiple fuzzy rules with the rule reduction. The thermodynamic deoxidation system is composed of several deaerators. When they work together the pressure of one deaerator is coupled with others and so is the water level. The pressure and the water level of each deaerator are coupled too. Thus the system is characterized with intensive coupling, multi-input multi-output (MIMO), great delay, time-varying and nonlinearity. Hence the precise mathematic model is difficult to obtain. Fuzzy controller can be implemented on the experience of experts without knowing the model. The number of rules is very large, but it can be reduced with the rule reduction of rough sets. The simulation investigation made for the design shows better robustness of the controller. The design of the fuzzy-rough controller proposed in this paper is simple, easy to achieve and suitable to other MIMO fuzzy controllers.