Eliminating Inrush Current in Three-Phase Transformer using Artificial Neural Network

Transformers are important parts of an electrical power system. When a power transformer is connected to the grid, usually inrush current increases substantially with a high value of harmonic components with a duration up to many cycles. The amount of flux in the core increases causing the magnetic...

Full description

Saved in:
Bibliographic Details
Main Author: Hussein A. Taha
Format: Article
Language:English
Published: Wasit University 2024-12-01
Series:Wasit Journal of Engineering Sciences
Subjects:
Online Access:https://ejuow.uowasit.edu.iq/index.php/ejuow/article/view/568
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Transformers are important parts of an electrical power system. When a power transformer is connected to the grid, usually inrush current increases substantially with a high value of harmonic components with a duration up to many cycles. The amount of flux in the core increases causing the magnetic circuit to saturate due to the increasing in the load. This paper describes a technique to accurately predict the inrush current and third harmonic of three phase transformer. A shallow neural network was created. The input parameters of the artificial neural network were the magnetization resistance Rm, the initial flux of phase A and the switching angle q. The number of neurons has been changed in the code to see the best performance value. The best validation performance was at epoch 71 with a value of 5.3641e-05. A good prediction results were obtained using this ANN. The simulation of the inrush current was done using the MATLAB Simulink software.  
ISSN:2305-6932
2663-1970