The determination and realization of compensation degrees in power control systems having load asymmetry by using fuzzy logic controller supported by artificial neural networks

Studies on power factor correction have been carried out for many years. But in these studies, phase powers received from supply are generally assumed to be balanced. This paper considers power control systems having unbalanced load. The fuzzy logic controller based on the fuzzy set theory provides a useful tool for converting the linguistic control from the heuristic or mathematical strategies; complex processes can be controlled in many situations effectively. But the most important and difficult point is how to obtain the proper control rules for a given system in the matter of fuzzy logic and artificial neural networks. ANNs have nonlinear structure and this gives an effective feature that approaches the results of learning phase well. Then, test phase gives results according to the input variable values in short time (degree about 10/sup -3/ second). It is the preferable method. In this paper, power factor correction is corrected to demand value by using fuzzy logic controller under load changing for the electric power control system containing current unbalance i.e. load asymmetry. In the control system, reference reactive power value is determined by ANN at various times of day as load changing with rising or falling trends. Consequently, the software of the system´s simulation results was given for power factor correction in the power control systems having load asymmetry. In addition to that, the developed method is compared with conventional compensation systems.