No abstract

An artificial neural network (ANN) based current controller for a high voltage direct current (HVDC) transmission link, composed of an ANN trained off-line in parallel with a robust PI controller, is described in this paper. Different ANN architectures are investigated for this ANN controller. Comparisons between the responses obtained with the PI and ANN controllers for the rectifier of a HVDC transmission system are made for various system ANN parameters (learning rate and momentum term) contingencies and it is shown that the later has many attractive features.

No abstract