A Contextual Anomaly Detection Framework for Energy Smart Meter Data Stream

Xiufeng Liu*, Zhichen Lai, Xin Wang, Lizhen Huang, Per Sieverts Nielsen

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

120 Downloads (Pure)

Abstract

Monitoring abnormal energy consumption is helpful for demand-side management. This paper proposes a framework for contextual anomaly detection (CAD) for residential energy consumption. This framework uses a sliding window approach and prediction-based detection method, along with the use of a concept drift method to identify the unusual energy consumption in different contextual environments. The anomalies are determined by a statistical method with a given threshold value. The paper evaluates the framework comprehensively using a real-world data set, compares with other methods and demonstrates the effectiveness and superiority.
Original languageEnglish
Title of host publicationProceedings of the International Conference on Neural Information Processing
PublisherSpringer
Publication date2020
Pages733-742
ISBN (Print)978-3-030-63822-1
DOIs
Publication statusPublished - 2020
EventInternational Conference on Neural Information Processing (ICONIP 2020) - Bankok, Thailand
Duration: 18 Nov 202022 Nov 2020

Conference

ConferenceInternational Conference on Neural Information Processing (ICONIP 2020)
Country/TerritoryThailand
CityBankok
Period18/11/202022/11/2020
SeriesNeural Information Processing - Letters and Reviews
Volume1333
ISSN1738-2572

Fingerprint

Dive into the research topics of 'A Contextual Anomaly Detection Framework for Energy Smart Meter Data Stream'. Together they form a unique fingerprint.

Cite this