Statistical identification and classification of potential islanding precursors in a grid-connected Solar Photo Voltaic system

Shashank Vyas, Rajesh Kumar, Rajesh Kavasseri

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

4 Citations (Scopus)

Abstract

Impacts of unintentional islanding will be more pronounced in the scenario of rising distributed-generation penetration especially on distribution grids. Solar photo voltaic generation is the forerunner technology whose grid-integration is increasing significantly. This paper explores unique anomalies occurring in a solar photo voltaic system integrated with a distribution network that can potentially cause unintentional islanding and applies statistical and machine learning techniques to identify them in real time. Grid-side disturbances and complete power-balance between the inverter and loads are the conditions deemed to occur frequently in a high penetration case. These have been simulated on a modified IEEE feeder and the resulting anomalies cross-verified in emulation and real-time simulation. Multivariate statistical techniques and a support vector classifier have been applied to identify the explored anomalous currents. The classification results show promising application towards a new predictive islanding detection logic inside solar inverters however the actual relay delay in tripping a breaker and the inverter's capability to trip before it on sensing a potential precursor to accidental islanding needs to be tested in real time.

Original languageEnglish
Title of host publication2016 IEEE 6th International Conference on Power Systems, ICPS 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781509001286
DOIs
Publication statusPublished - 5 Oct 2016
Externally publishedYes
Event6th IEEE International Conference on Power Systems, ICPS 2016 - New Delhi, India
Duration: 4 Mar 20166 Mar 2016

Publication series

Name2016 IEEE 6th International Conference on Power Systems, ICPS 2016

Conference

Conference6th IEEE International Conference on Power Systems, ICPS 2016
Country/TerritoryIndia
CityNew Delhi
Period4/03/166/03/16

Keywords

  • Classification
  • Inverters
  • Islanding
  • Machine learning
  • Power system

ASJC Scopus subject areas

  • Energy Engineering and Power Technology
  • Renewable Energy, Sustainability and the Environment
  • Electrical and Electronic Engineering
  • Control and Optimization

Fingerprint

Dive into the research topics of 'Statistical identification and classification of potential islanding precursors in a grid-connected Solar Photo Voltaic system'. Together they form a unique fingerprint.

Cite this