Online Anomaly Energy Consumption Detection Using Lambda Architecture

Xiufeng Liu, Nadeem Iftikhar, Per Sieverts Nielsen, Alfred Heller

Research output: Contribution to journalJournal articleResearchpeer-review

936 Downloads (Pure)

Abstract

With the widely use of smart meters in the energy sector, anomaly detection becomes a crucial mean to study the unusual consumption behaviors of customers, and to discover unexpected events of using energy promptly. Detecting consumption anomalies is, essentially, a real-time big data analytics problem, which does data mining on a large amount of parallel data streams from smart meters. In this paper, we propose a supervised learning and statistical-based anomaly detection method, and implement a Lambda system using the in-memory distributed computing framework, Spark and its extension Spark Streaming. The system supports not only iterative refreshing the detection models from scalable data sets, but also real-time anomaly detection on scalable live data streams. This paper empirically evaluates the system and the detection algorithm, and the results show the effectiveness and the scalability of the lambda detection system.
Original languageEnglish
Book seriesLecture Notes in Computer Science
Pages (from-to)193-209
Number of pages17
ISSN0302-9743
DOIs
Publication statusPublished - 2016
Event18th International Conference on Big Data Analytics and Knowledge Discovery (DaWaK 2016) - Porto, Portugal
Duration: 5 Sep 20168 Sep 2016
Conference number: 18
http://www.dexa.org/dawak2016

Conference

Conference18th International Conference on Big Data Analytics and Knowledge Discovery (DaWaK 2016)
Number18
CountryPortugal
CityPorto
Period05/09/201608/09/2016
Internet address

Keywords

  • Anomaly detection
  • Real-time
  • Lambda architecture
  • Data mining

Cite this

@article{56891e1289454ef99347086fbcac2698,
title = "Online Anomaly Energy Consumption Detection Using Lambda Architecture",
abstract = "With the widely use of smart meters in the energy sector, anomaly detection becomes a crucial mean to study the unusual consumption behaviors of customers, and to discover unexpected events of using energy promptly. Detecting consumption anomalies is, essentially, a real-time big data analytics problem, which does data mining on a large amount of parallel data streams from smart meters. In this paper, we propose a supervised learning and statistical-based anomaly detection method, and implement a Lambda system using the in-memory distributed computing framework, Spark and its extension Spark Streaming. The system supports not only iterative refreshing the detection models from scalable data sets, but also real-time anomaly detection on scalable live data streams. This paper empirically evaluates the system and the detection algorithm, and the results show the effectiveness and the scalability of the lambda detection system.",
keywords = "Anomaly detection, Real-time, Lambda architecture, Data mining",
author = "Xiufeng Liu and Nadeem Iftikhar and Nielsen, {Per Sieverts} and Alfred Heller",
year = "2016",
doi = "10.1007/978-3-319-43946-4_13",
language = "English",
pages = "193--209",
journal = "Lecture Notes in Computer Science",
issn = "0302-9743",
publisher = "Springer",

}

Online Anomaly Energy Consumption Detection Using Lambda Architecture. / Liu, Xiufeng; Iftikhar, Nadeem ; Nielsen, Per Sieverts; Heller, Alfred.

In: Lecture Notes in Computer Science, 2016, p. 193-209.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Online Anomaly Energy Consumption Detection Using Lambda Architecture

AU - Liu, Xiufeng

AU - Iftikhar, Nadeem

AU - Nielsen, Per Sieverts

AU - Heller, Alfred

PY - 2016

Y1 - 2016

N2 - With the widely use of smart meters in the energy sector, anomaly detection becomes a crucial mean to study the unusual consumption behaviors of customers, and to discover unexpected events of using energy promptly. Detecting consumption anomalies is, essentially, a real-time big data analytics problem, which does data mining on a large amount of parallel data streams from smart meters. In this paper, we propose a supervised learning and statistical-based anomaly detection method, and implement a Lambda system using the in-memory distributed computing framework, Spark and its extension Spark Streaming. The system supports not only iterative refreshing the detection models from scalable data sets, but also real-time anomaly detection on scalable live data streams. This paper empirically evaluates the system and the detection algorithm, and the results show the effectiveness and the scalability of the lambda detection system.

AB - With the widely use of smart meters in the energy sector, anomaly detection becomes a crucial mean to study the unusual consumption behaviors of customers, and to discover unexpected events of using energy promptly. Detecting consumption anomalies is, essentially, a real-time big data analytics problem, which does data mining on a large amount of parallel data streams from smart meters. In this paper, we propose a supervised learning and statistical-based anomaly detection method, and implement a Lambda system using the in-memory distributed computing framework, Spark and its extension Spark Streaming. The system supports not only iterative refreshing the detection models from scalable data sets, but also real-time anomaly detection on scalable live data streams. This paper empirically evaluates the system and the detection algorithm, and the results show the effectiveness and the scalability of the lambda detection system.

KW - Anomaly detection

KW - Real-time

KW - Lambda architecture

KW - Data mining

U2 - 10.1007/978-3-319-43946-4_13

DO - 10.1007/978-3-319-43946-4_13

M3 - Journal article

SP - 193

EP - 209

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

SN - 0302-9743

ER -