Neural Network based Adaptive Overcurrent Directional Protection in Large Electrical Power Networks

N.H. El-Amary, Y.G. Monir, and T. Youssif (Egypt)


Adaptive protection, directional over current relay, maximum sensitivity towards angel, neural network technique, electrical power system analysis software (ETAP)


Power system automation, nowadays, enriches all fields of power studies. In this paper, a directional over current relay is developed, utilizing neural network technique. Directional relays play an indispensable role for protection coordination in large power networks. The principal of operation of directional over current relays depend mainly on different combinations of phase/line currents and polarization voltage in order to inhibit relays malfunction and to assure prevention of false operation. The different combinations are mainly different connections of a specific phase/line current with a specific phase/line voltage. A new adaptive artificial neural network based relay is developed in order to adopt the maximum sensitivity torque angle and to define the optimum polarized current and voltage parameters to be considered in the relay operation. The new artificial developed system is tested on a large simulated unified electrical power network. The power network is simulated using electrical power system analysis software (ETAP). The developed system is capable of selecting the best current-voltage combination at different fault locations, types and grounding systems as well.

Important Links:

Go Back