Power Grids Partitioning by Detection of Electrical Communities

Liang Tian, Xiaoli Zhao, Yiran Wang, Lechuan Piao, Fei Xue, Qigang Wu

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

Abstract

In rapidly evolving power grids, effective partitioning strategies are critical for ensuring stability, reliability, and efficiency in subsystems and even the operation of the entire system. Community detection has been considered promising in power grids partition recently, but the electrical features were not well integrated with most conventional algorithms. This study proposes a novel Inversed Girvan-Newman Algorithm to partition power grids. Compared with traditional Girvan-Newman Algorithm, the updated electrical betweenness plays a totally different role in principle according to basic understanding of electrical community. Case studies in IEEE-lIS and IEEE-300 bus transmission networks are implemented to detect electrical communities, which has shown better performance in modularity, boundary flow, and subsystem operational cost.

Original languageEnglish
Title of host publicationProceedings of 2024 IEEE 7th International Electrical and Energy Conference, CIEEC 2024
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages155-160
Number of pages6
ISBN (Electronic)9798350359558
DOIs
Publication statusPublished - 2024
Event7th IEEE International Electrical and Energy Conference, CIEEC 2024 - Harbin, China
Duration: 10 May 202412 May 2024

Publication series

NameProceedings of 2024 IEEE 7th International Electrical and Energy Conference, CIEEC 2024

Conference

Conference7th IEEE International Electrical and Energy Conference, CIEEC 2024
Country/TerritoryChina
CityHarbin
Period10/05/2412/05/24

Keywords

  • community detection
  • Complex Network
  • Electrical Edge Betweenness
  • Girvan-Newman Algorithm
  • Modularity

Fingerprint

Dive into the research topics of 'Power Grids Partitioning by Detection of Electrical Communities'. Together they form a unique fingerprint.

Cite this