Automated Power Factor Improvement Based on Artificial Neural Networks

This paper addresses the basic issues related to voltage instability considering the characteristics of transmission systems and thereby introducing a power factor improvement tool based on artificial neural networks focusing on the compensation of reactive power, consequently improving the voltage stability. A two layered feed-forward neural network employing back propagation algorithm was used to train the IEEE 22 bus system on a MATLAB environment.The capacitance to be introduced is predicted by the NN fitting tool and the predicted capacitance is to be switched to the load lines. The load flow of the system before and after the introduction of the capacitance has been analyzed and verified on an ETAP environment.