Indirect tuning of membership function in a fixed fuzzy structure for efficient control of a DC drive system

A novel approach has been developed to tune the parameters of the membership function in a fixed fuzzy logic structure indirectly. The proposed concept is based on integrating the membership values of the input variables for a given linguistic set having a fixed centre but different widths. An arithmetical averaging method is applied to estimate the resultant membership value. This idea resembles that of combining the knowledge of different experts based on their independent assessment analyzing a given problem and then to derive a pseudo-optimal decision for the final action. This work is applied to evaluate the change of firing angle required to control the speed of a seperately excited DC motor drive fed from a phase-controlled converter. The response of the motor is found to be excellent in comparison to conventional fuzzy structure.