Time-series decomposition of power demand data to extract uncertain features

Tomoya Imanishi, Masahiro Yoshida, Janaka Wijekoon, Hiroaki Nishi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

The spread of smart meters means that a large amount of power demand information from private houses is being collected around the world. Owing to the development of smart city infrastructure, the use of standardized frameworks for extracting features from power demand information has become vital. In this paper, we propose a novel decomposition approach useful for extracting feature values from power demand information from a house. Energy consumption was monitored for multiple houses for one month in Japan with a sampling duration of 30 minutes, which is a standard sampling time of smart meters in Japan. First, periodic characteristics were detected for 24 hours based on autocorrelation analysis. Then, the monitored information was decomposed into four components: standby power, trends, and periodic and residual parts. The distribution of the residual part is similar to a Gaussian distribution, so the behavior of the residual part was parameterized using variance and average. Trend, periodic, and residual components were clustered by means of k-means clustering in order to aggregate the difference in behaviors. There was no periodic component in the residual part according to auto-correlation analysis. Nevertheless, some clusters had a relatively large variance, which means that abnormal power demand occurred frequently in datasets. The amount of variance and climate correlation was analyzed, and the fact detected that large scale events disturb usual daily life-styles, from the viewpoint of energy usage. Last, these features were compared with actual customer information. In the evaluation, family structure and floor space were utilized to prove the effectiveness of the proposed decomposition approach. The evaluation proved that this decomposition method could extract uncertainty features from power demand information.

Original languageEnglish
Title of host publicationProceedings - 2017 IEEE International Symposium on Industrial Electronics, ISIE 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1535-1540
Number of pages6
ISBN (Electronic)9781509014125
DOIs
Publication statusPublished - 2017 Aug 3
Event26th IEEE International Symposium on Industrial Electronics, ISIE 2017 - Edinburgh, Scotland, United Kingdom
Duration: 2017 Jun 182017 Jun 21

Other

Other26th IEEE International Symposium on Industrial Electronics, ISIE 2017
CountryUnited Kingdom
CityEdinburgh, Scotland
Period17/6/1817/6/21

Fingerprint

Smart meters
Time series
Decomposition
Autocorrelation
Sampling
Gaussian distribution
Energy utilization
Uncertainty
Smart city

Keywords

  • Classification
  • Discrete fourier transform
  • Feature extraction
  • Power demand information
  • Support vector machine

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Control and Systems Engineering

Cite this

Imanishi, T., Yoshida, M., Wijekoon, J., & Nishi, H. (2017). Time-series decomposition of power demand data to extract uncertain features. In Proceedings - 2017 IEEE International Symposium on Industrial Electronics, ISIE 2017 (pp. 1535-1540). [8001473] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ISIE.2017.8001473

Time-series decomposition of power demand data to extract uncertain features. / Imanishi, Tomoya; Yoshida, Masahiro; Wijekoon, Janaka; Nishi, Hiroaki.

Proceedings - 2017 IEEE International Symposium on Industrial Electronics, ISIE 2017. Institute of Electrical and Electronics Engineers Inc., 2017. p. 1535-1540 8001473.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Imanishi, T, Yoshida, M, Wijekoon, J & Nishi, H 2017, Time-series decomposition of power demand data to extract uncertain features. in Proceedings - 2017 IEEE International Symposium on Industrial Electronics, ISIE 2017., 8001473, Institute of Electrical and Electronics Engineers Inc., pp. 1535-1540, 26th IEEE International Symposium on Industrial Electronics, ISIE 2017, Edinburgh, Scotland, United Kingdom, 17/6/18. https://doi.org/10.1109/ISIE.2017.8001473
Imanishi T, Yoshida M, Wijekoon J, Nishi H. Time-series decomposition of power demand data to extract uncertain features. In Proceedings - 2017 IEEE International Symposium on Industrial Electronics, ISIE 2017. Institute of Electrical and Electronics Engineers Inc. 2017. p. 1535-1540. 8001473 https://doi.org/10.1109/ISIE.2017.8001473
Imanishi, Tomoya ; Yoshida, Masahiro ; Wijekoon, Janaka ; Nishi, Hiroaki. / Time-series decomposition of power demand data to extract uncertain features. Proceedings - 2017 IEEE International Symposium on Industrial Electronics, ISIE 2017. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 1535-1540
@inproceedings{fd80df6ef3934677a0788e946787807a,
title = "Time-series decomposition of power demand data to extract uncertain features",
abstract = "The spread of smart meters means that a large amount of power demand information from private houses is being collected around the world. Owing to the development of smart city infrastructure, the use of standardized frameworks for extracting features from power demand information has become vital. In this paper, we propose a novel decomposition approach useful for extracting feature values from power demand information from a house. Energy consumption was monitored for multiple houses for one month in Japan with a sampling duration of 30 minutes, which is a standard sampling time of smart meters in Japan. First, periodic characteristics were detected for 24 hours based on autocorrelation analysis. Then, the monitored information was decomposed into four components: standby power, trends, and periodic and residual parts. The distribution of the residual part is similar to a Gaussian distribution, so the behavior of the residual part was parameterized using variance and average. Trend, periodic, and residual components were clustered by means of k-means clustering in order to aggregate the difference in behaviors. There was no periodic component in the residual part according to auto-correlation analysis. Nevertheless, some clusters had a relatively large variance, which means that abnormal power demand occurred frequently in datasets. The amount of variance and climate correlation was analyzed, and the fact detected that large scale events disturb usual daily life-styles, from the viewpoint of energy usage. Last, these features were compared with actual customer information. In the evaluation, family structure and floor space were utilized to prove the effectiveness of the proposed decomposition approach. The evaluation proved that this decomposition method could extract uncertainty features from power demand information.",
keywords = "Classification, Discrete fourier transform, Feature extraction, Power demand information, Support vector machine",
author = "Tomoya Imanishi and Masahiro Yoshida and Janaka Wijekoon and Hiroaki Nishi",
year = "2017",
month = "8",
day = "3",
doi = "10.1109/ISIE.2017.8001473",
language = "English",
pages = "1535--1540",
booktitle = "Proceedings - 2017 IEEE International Symposium on Industrial Electronics, ISIE 2017",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Time-series decomposition of power demand data to extract uncertain features

AU - Imanishi, Tomoya

AU - Yoshida, Masahiro

AU - Wijekoon, Janaka

AU - Nishi, Hiroaki

PY - 2017/8/3

Y1 - 2017/8/3

N2 - The spread of smart meters means that a large amount of power demand information from private houses is being collected around the world. Owing to the development of smart city infrastructure, the use of standardized frameworks for extracting features from power demand information has become vital. In this paper, we propose a novel decomposition approach useful for extracting feature values from power demand information from a house. Energy consumption was monitored for multiple houses for one month in Japan with a sampling duration of 30 minutes, which is a standard sampling time of smart meters in Japan. First, periodic characteristics were detected for 24 hours based on autocorrelation analysis. Then, the monitored information was decomposed into four components: standby power, trends, and periodic and residual parts. The distribution of the residual part is similar to a Gaussian distribution, so the behavior of the residual part was parameterized using variance and average. Trend, periodic, and residual components were clustered by means of k-means clustering in order to aggregate the difference in behaviors. There was no periodic component in the residual part according to auto-correlation analysis. Nevertheless, some clusters had a relatively large variance, which means that abnormal power demand occurred frequently in datasets. The amount of variance and climate correlation was analyzed, and the fact detected that large scale events disturb usual daily life-styles, from the viewpoint of energy usage. Last, these features were compared with actual customer information. In the evaluation, family structure and floor space were utilized to prove the effectiveness of the proposed decomposition approach. The evaluation proved that this decomposition method could extract uncertainty features from power demand information.

AB - The spread of smart meters means that a large amount of power demand information from private houses is being collected around the world. Owing to the development of smart city infrastructure, the use of standardized frameworks for extracting features from power demand information has become vital. In this paper, we propose a novel decomposition approach useful for extracting feature values from power demand information from a house. Energy consumption was monitored for multiple houses for one month in Japan with a sampling duration of 30 minutes, which is a standard sampling time of smart meters in Japan. First, periodic characteristics were detected for 24 hours based on autocorrelation analysis. Then, the monitored information was decomposed into four components: standby power, trends, and periodic and residual parts. The distribution of the residual part is similar to a Gaussian distribution, so the behavior of the residual part was parameterized using variance and average. Trend, periodic, and residual components were clustered by means of k-means clustering in order to aggregate the difference in behaviors. There was no periodic component in the residual part according to auto-correlation analysis. Nevertheless, some clusters had a relatively large variance, which means that abnormal power demand occurred frequently in datasets. The amount of variance and climate correlation was analyzed, and the fact detected that large scale events disturb usual daily life-styles, from the viewpoint of energy usage. Last, these features were compared with actual customer information. In the evaluation, family structure and floor space were utilized to prove the effectiveness of the proposed decomposition approach. The evaluation proved that this decomposition method could extract uncertainty features from power demand information.

KW - Classification

KW - Discrete fourier transform

KW - Feature extraction

KW - Power demand information

KW - Support vector machine

UR - http://www.scopus.com/inward/record.url?scp=85029894034&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029894034&partnerID=8YFLogxK

U2 - 10.1109/ISIE.2017.8001473

DO - 10.1109/ISIE.2017.8001473

M3 - Conference contribution

SP - 1535

EP - 1540

BT - Proceedings - 2017 IEEE International Symposium on Industrial Electronics, ISIE 2017

PB - Institute of Electrical and Electronics Engineers Inc.

ER -