Smart non-active power compensator with neural network control-an evaluation

Distortion in power network nonactive power components can be eliminated by a compensator that consists of one or a few power converters. To obtain cost-optimal compensation, a specific distortion should be eliminated by a compensator configuration with the lowest operational costs. In this paper, a multi-converter compensator is proposed with neural network control and investigated by simulation. The neural network control system analyses the distortion components-nonactive power and the losses in the power converters of the compensator-and chooses this converter (or converters) and their working mode that ensures cost-optimal compensation