Application of neural networks to control a synchronous generator based on
direct adaptive control scheme is investigated in this paper. Use of a neur
al network to model the dynamic system is avoided by making use of the sign
of the Jacobian of the plant. This substantially reduces the complexity an
d the computation time of the control algorithm. The controller is trained
on-line using the back-propagation algorithm which gives an adaptive attrib
ute to it. Moreover, the controller does not need the state information and
only employs the plant outputs. Simulation results are presented to comple
ment the theoretical discussion.