High impedance fault detection method efficiency: Simulation vs. real-world data acquisition
Authors:
- 1 February 2015
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Abstract
In this paper, the reliability of a novel high impedance fault (HIF) detection method is compared for two data acquisition cases, namely, real-world and simulation. The real-world data is collected in high current research laboratory, and the simulation data is acquired using PSCAD/EMTDC software. The real-world condition is set to the most practical conditions (noisy environment, magnetic-core CTs). It is observed that although the proposed HIF detection method shows perfect reliability for the simulated HIF, its reliability for real-world data is considerably lower. From the comparison between the proposed detection algorithm reliability between the simulation and real-world data, it is concluded that it is not sufficient to prove the method reliability solely using the data acquired from simulation or impractical real-world cases, measurement noise. In order to increase the method reliability for real-world data, four actions are proposed in this paper; using a more complex classifier, a larger feature set, a wavelet based de-noising algorithm, and decreasing the method speed.Keywords
This publication has 19 references indexed in Scilit:
- Active fault location in distribution network using time-frequency reflectometry2015
- A novel islanding detection method for constant current inverter based distributed generations2011
- Investigation of influential factors on passive islanding detection methods of inverter based distributed generation2011
- Power Quality Indices for Transient DisturbancesIEEE Transactions on Power Delivery, 2005
- High Impedance Fault Detection Based on Wavelet Transform and Statistical Pattern RecognitionIEEE Transactions on Power Delivery, 2005
- High breakdown estimators for principal components: the projection-pursuit approach revisitedJournal of Multivariate Analysis, 2005
- Classification of Power Quality Events Using Optimal Time-Frequency Representations—Part 1: TheoryIEEE Transactions on Power Delivery, 2004
- Recognizing human actions: a local SVM approach2004
- An implementation of a hybrid intelligent tool for distribution system fault diagnosisIEEE Transactions on Power Delivery, 1997
- A survey of moment-based techniques for unoccluded object representation and recognitionCVGIP: Graphical Models and Image Processing, 1992