Development of an artificial neural network based relay to detect islanding condition of embedded generation

S. K. Salman, David J. King

Research output: Chapter in Book/Report/Conference proceedingConference contribution


With the increase of number of embedded generators (EGs) in public utility networks the need for adequate protection schemes has become extremely important. In particular, safety problems that arise due to loss of mains (LOM), and the inability of current protection, under certain circumstances, to detect it means that there is a lot of current research into possible alternative methods for detecting LOM. Also, the use of artificial intelligence (AI) in power system operation is becoming widespread. In particular, artificial neural networks (ANN) are being developed to tackle a number of power system control and protection problems. One of the most important things to consider when using ANN is deciding upon appropriate and relevant training/testing data. This paper results from a project to find a new method of detection of LOM for EG using ANN, and outlines the methods used to develop suitable training/testing data
Original languageEnglish
Title of host publicationProceedings of the 6th International Power Engineering Conference
Place of PublicationSingapore
PublisherNanyang Technological University
Number of pages6
ISBN (Print)9810487908
Publication statusPublished - 2003
Event6th International Power Engineering Conference - Singapore, Singapore
Duration: 27 Nov 200329 Nov 2003
Conference number: 6


Conference6th International Power Engineering Conference
Abbreviated titleIPEC 2003


  • Embedded generation
  • Loss of mains protection
  • Islanding
  • Modelling
  • EMTP


Dive into the research topics of 'Development of an artificial neural network based relay to detect islanding condition of embedded generation'. Together they form a unique fingerprint.

Cite this