International Journal of Power Electronics and Drive System (IJPEDS) Vol. 11, No. 1, March 2020, pp. 505~514 ISSN: 2088-8694, DOI: 10.11591/ijpeds.v11.i1.pp505-514  505 Journal homepage: http://ijpeds.iaescore.com Improvement of protection relay with a single phase auto- reclosing mechanism based on artificial neural network Zozan Saadallah Hussain 1 , Ahmed J. Ali 2 , Ahmed A. Allu 3 , Rakan Khalil Antar 4 1,3 Technical Institute, Northern Technical University, Iraq 2,4 Technical College, Northern Technical University, Iraq Article Info ABSTRACT Article history: Received Mar 26, 2019 Revised Jul 8, 2019 Accepted Oct 22, 2019 This paper presents a developed logical tripping scheme to improve conventional protection performance. Adaptive single pole auto-reclosure (ASPAR) system is proposed that considers, automatically tripping and reclosing of a multi-shot independent pole technique of a circuit breaker at a predetermined sequence, which can be used to boost the synchronization of the power grid under the transient fault conditions. Moreover, the ASPAR can be utilized to enhance the electrical system stability and reliability at the same operating conditions. Based on the three-phase system, the Artificial neural network (ANN) in this work has been done in order to diagnose and detect healthy and faulted phases. The proposed ANN-fault classifier method consists of the logic gates, router circuits, timers, and positive and negative sequence analyses circuit. In addition, it is used to give the ability to recognize a fault type, which by training on the sequence angle values and coordination of the transmission line. Three-phase overhead transmission line including the proposed ASPAR is built in MATLAB\SIMULINK environment. Thus the performance ANN-fault classified is tested under different fault conditions. Simulation results show that the proposed ASPAR based on ANN is accurate and well performance. Whereas resultant tripping and reclosing signals of ASPAR are successfully provided that enhances the circuit breaker mechanism under these operating condition. Keywords: Artificial neural network Fault classification Matlab\Simulink prog. Power system Single phase auto-reclosure This is an open access article under the CC BY-SA license. Corresponding Author: Zozan Saadallah Hussain, Technical Institute, Northern Technical University, Mosul, Iraq, Email: zozan.hussian@gmail.com 1. INTRODUCTION Detection and classification of faults occurred at transmission lines are considered an important factor in correctly, safely and reliable operation of protection relays. Intelligent technologies had been widely used in various areas of electrical power applications that require a consciously monitoring to ensure the reliability of the system to equip consumers with uninterrupted power [1]. Different types of these methods have been handled to detect and identify the transient and permanent faults that occur because of external influences such as natural disasters or due to increase loads more than the capacity of the generation system. A fuzzy systems technique was used to detect and classify faults because of it has a great ability to analyze the changes that occur in the elements of the power system at the instant of occurring failure [2-3]. However, this technique suffers from the problem of selecting the type of membership functions of the controller, which had been solved by using intelligent neural networks. It is function is to decide and choose the shape, number, and type of these functions [4].