A Novel Approach for Discrimination Magnetizing Inrush Current and Internal Fault in Power Transformers Based on Neural Network

Subject Areas : B. Computer Systems Organization

Mehran Taghipour-Gorjikolaie ¹ (Department of Electrical and Computer Engineering, University of Birjand, Birjand ,Iran)
Mohammad Yazdani-Asrami ² (Department of Electrical and Computer Engineering, Babol University of Technology, Babol ,Iran)
S. Asghar Gholamian ³ (Department of Electrical and Computer Engineering, Babol University of Technology, Babol ,Iran)
S. Mohammad Razavi ⁴ (Department of Electrical and Computer Engineering, University of Birjand, Birjand ,Iran)

Received: 2015-08-31 Accepted : 2015-08-31 Published : 2015-08-01

Keywords: Artificial neural network, Activation Function, differential protection, Magnetizing Inrush Current, Improved Gravitational Search Algorithm, and Transformer Fault,

Abstract :

One of the major problems that may occur in the differential protection systems of power transformers is mal-operation of the protection relays in sake of internal fault detection, because of similarity between this current and inrush current. This paper presents a novel approach for discriminating inrush current from internal fault in power transformers based on Improved Gravitational Search Algorithm (IGSA). For this purpose, an Artificial Neural Network (ANN) which is trained by IGSA has been applied to discrete sample data of internal fault and inrush currents in the transformers. Results show that, the used approach can discriminate between these two kinds of phenomenon, very well and also, has high accuracy and excellent reliability, in addition, it has less computational burden and complexity.