This paper is concerned with the application of an adaptive, fuzzy log
ic controller to a synchronous generator system. Significant reduction
s in the amount of required calculations have been made which allow a
direct analysis of the system. A novel on-line training algorithm is p
roposed to adapt fuzzy rule base which is shown to converge globally a
nd provide the desired regulation for disturbances at all operating po
ints, The sensitivity of the controller with respect to changes in des
ign parameters is analysed. Test responses are compared with a self-tu
ning regulator to show the effectiveness of the proposed system. Micro
machine tests further establish the possibility of developing a practi
cal controller.