An intelligent hybrid active power factor compensator for power systems

The power factor, in present days, is a critical matter. The ever increasing utilization of power electronics in all kinds of industry applications and the severe standards requirements, are pushing the research toward new solutions to keep the industrial power factor around certain values. The switched capacitors banks brings together problems that are not easy to solve, like: parallel resonance and over voltage. This article describes the application of a hybrid converter as a line conditioner. The hybrid converter basically is a converter bridge with two GTOs, that transform it in a fully controllable inverter. Therefore, this converter allows independent control of voltage and displacement factor. When appropriately controlled this converter can produce part of the reactive power consumed by the load. Made with a well known technology, this converter is suitable for large power compensation. Theoretical analysis and experimental results are presented. Also, a control strategy based on learning is proposed. This structure, a hierarchical approach, uses hints, coded in fuzzy conditional rules, to train a neural network responsible for the converter variables manipulation