Evaluation of islanding detection techniques for inverter-based distributed generation

Omar N. Faqhruldin, E. F. El-Saadany, H. H. Zeineldin

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Scopus citations

Abstract

In this paper; four islanding detection techniques for inverter-based distributed generator (DG) are presented. The techniques are: decision tree (DT), support vector machine (SVM), radial basis function network (RBF), and probabilistic neural network (PNN). In literature, these techniques were proposed as islanding detection methods. However, the proposed techniques face various limitations such as the size and type of the used distribution network and the limitation of the extracted features. This paper overcomes these limitations and gives a very accurate comparison between these techniques by extracting seven features from damped-sinusoid model of the voltage and frequency waveforms using the MATLAB/SIMULINK and also using the IEEE 34-bus distribution system. The results show that out of the four tested techniques, PNN technique can accurately detect islanding for inverter based DG.

Original languageBritish English
Title of host publication2012 IEEE Power and Energy Society General Meeting, PES 2012
DOIs
StatePublished - 2012
Event2012 IEEE Power and Energy Society General Meeting, PES 2012 - San Diego, CA, United States
Duration: 22 Jul 201226 Jul 2012

Publication series

NameIEEE Power and Energy Society General Meeting
ISSN (Print)1944-9925
ISSN (Electronic)1944-9933

Conference

Conference2012 IEEE Power and Energy Society General Meeting, PES 2012
Country/TerritoryUnited States
CitySan Diego, CA
Period22/07/1226/07/12

Keywords

  • decision tree
  • inverter-based distributed generator
  • islanding detection
  • power systems
  • probabilistic neural network
  • radial basis function neural network
  • support vector machine

Fingerprint

Dive into the research topics of 'Evaluation of islanding detection techniques for inverter-based distributed generation'. Together they form a unique fingerprint.

Cite this