Reduced rule fuzzy logic controller for performance improvement of process control

The shortcomings in conventional controllers have diverted the attention of the researchers to concentrate on the use of artificial intelligent controllers such as fuzzy logic controllers. The fuzzy controllers are intuitive and give better response compared to conventional controllers. The performance of these controllers can be improved by increasing the rules present in the rule base. With increased rules the process slows down thereby reducing the effectiveness of the controller. In this paper, a novel method is proposed to reduce the number of rules that are fired while keeping the performance similar to that of large rules. The proposed method utilizes the membership value of the linguistic variable to calculate equilibrium value and the rules are fired only if both the inputs have membership value higher than the equilibrium value. By utilizing this method the rules that are fired are reduced from 49 to 16 rules. With the reduction of fuzzy rules the computational memory and computational time required are reduced considerably. The proposed fuzzy controller is applied to a second order system and non-linear system. Simulation results are presented and analyzed to validate the proposed method.