Archives


Rate this Article: (0) Votes

Research Article

Year : 2019 | Volume: 3 | Issue: 1 | Pages: 14-23

k-Nearest Neighbor based Classification Algorithm for Phase to Phase Faults in 6-Phase Transmission Lines

A NARESH KUMAR1*, 2

doi:10.24951/sreyasijst.org/2019011003

Corresponding author

A NARESH KUMAR*

Department of Electrical and Electronics Engineering, Institute of Aeronautical Engineering

  • 1. Department of Electrical and Electronics Engineering, Institute of Aeronautical Engineering

Received on: 1/6/2019

Revised on: 1/6/2019

Accepted on: 1/6/2019

Published on: 1/6/2019

Abstract

This paper presents a method for classification of phase to phase faults in six phase transmission (SPT) line using k-nearest neighbor (k-NN) approach. A 68 km length, 138 kV SPT line has been simulated using Simulink® and Simpowersystems® toolbox of MATLAB software. The fundamental component of currents measured at source location is given as input to k-NN system. The impact of various fault distance location and fault inception angle has been validated on the performance of k-NN protection approach. The results of k-NN based fault classifier show that it correctly classified all types of phase to phase faults within half cycle. It verifies the accuracy of k-NN approach.

Keywords

Fault Classification, k-Nearest Neighbor, phase to phase faults, Six Phase Transmission Lines,k-NN approach.