KNN-ADMM based Online Missing Data Recovery in Electricity Markets

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

1 Scopus citations

Abstract

This paper proposes a missing electricity markets data online recovery strategy combining the improved K-Nearest Neighbor clustering(KNN) and Alternating Direction Method of Multipliers(ADMM). Considering the coupling consumption behaviors in different end users, the electricity markets data can be approximated in low-rank feature with singular value decomposition method. Moreover, the sliding time window is adopted as an online repair strategy to realize real-time recovery and improve the speed of recovery. Ultimately, the effectiveness of proposed online KNN-ADMM algorithm is verified through the electricity markets data of four Chinese cities. Results show that the algorithm is suitable for electricity markets data online recovery and the recovery error is within 5% with the 15% missing data.

Original languageEnglish
Title of host publicationProceedings of the 34th Chinese Control and Decision Conference, CCDC 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages4030-4035
Number of pages6
ISBN (Electronic)9781665478960
DOIs
StatePublished - 2022
Event34th Chinese Control and Decision Conference, CCDC 2022 - Hefei, China
Duration: 15 Aug 202217 Aug 2022

Publication series

NameProceedings of the 34th Chinese Control and Decision Conference, CCDC 2022

Conference

Conference34th Chinese Control and Decision Conference, CCDC 2022
Country/TerritoryChina
CityHefei
Period15/08/2217/08/22

Keywords

  • Alternating Direction Method of Multipliers(ADMM)
  • Data Recovery
  • Electricity Markets Data
  • K-Nearest Neighbor(KNN)

Fingerprint

Dive into the research topics of 'KNN-ADMM based Online Missing Data Recovery in Electricity Markets'. Together they form a unique fingerprint.

Cite this