Advanced search
1 file | 1.43 MB Add to list

Scheduled maintenance: Publication downloads temporarily unavailable.

Due to maintenance publication downloads will not be available on:

  • Wednesday, March 27, 17:00 – 21:00
  • Thursday, March 28, 17:00 – 21:00

Exports of lists, FWO and BOF information will remain available.

For any questions, please contact biblio@ugent.be. Apologies for any inconveniences, and thank you for your understanding.

Two-stage load pattern clustering using fast wavelet transformation

Kevin Mets (UGent) , Frederick Depuydt (UGent) and Chris Develder (UGent)
(2016) IEEE TRANSACTIONS ON SMART GRID. 7(5). p.2250-2259
Author
Organization
Abstract
Smart grids collect large volumes of smart meter data in the form of time series, or so-called load patterns. We outline the applications that benefit from analyzing this data (ranging from customer segmentation to operational system planning), and propose two-stage load pattern clustering. The first stage is performed per individual user and identifies the various typical daily power usage patterns (s) he exhibits. The second stage takes those typical user patterns as input to group users that are similar. To improve scalability, we use fast wavelet transformation (FWT) of the time series data, which reduces the dimensionality of the feature space where the clustering algorithm operates (i.e., from N data points in the time domain to log N). Another qualitative benefit of FWT is that patterns that are identical in shape, but just differ in a (typically small) time shift still end up in the same cluster. Furthermore, we use g-means instead of k-means as the clustering algorithm. Our comprehensive set of experiments analyzes the impact of using FWT versus time-domain features, and g-versus k-means, to conclude that in terms of cluster quality metrics our system is comparable to state-of-the-art methods, while being more scalable (because of the dimensionality reduction).
Keywords
IBCN, ELECTRICITY CUSTOMERS, CLASSIFICATION, PROFILES, IDENTIFICATION, SEGMENTATION, CONSUMERS, TIME

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 1.43 MB

Citation

Please use this url to cite or link to this publication:

MLA
Mets, Kevin, et al. “Two-Stage Load Pattern Clustering Using Fast Wavelet Transformation.” IEEE TRANSACTIONS ON SMART GRID, vol. 7, no. 5, 2016, pp. 2250–59, doi:10.1109/TSG.2015.2446935.
APA
Mets, K., Depuydt, F., & Develder, C. (2016). Two-stage load pattern clustering using fast wavelet transformation. IEEE TRANSACTIONS ON SMART GRID, 7(5), 2250–2259. https://doi.org/10.1109/TSG.2015.2446935
Chicago author-date
Mets, Kevin, Frederick Depuydt, and Chris Develder. 2016. “Two-Stage Load Pattern Clustering Using Fast Wavelet Transformation.” IEEE TRANSACTIONS ON SMART GRID 7 (5): 2250–59. https://doi.org/10.1109/TSG.2015.2446935.
Chicago author-date (all authors)
Mets, Kevin, Frederick Depuydt, and Chris Develder. 2016. “Two-Stage Load Pattern Clustering Using Fast Wavelet Transformation.” IEEE TRANSACTIONS ON SMART GRID 7 (5): 2250–2259. doi:10.1109/TSG.2015.2446935.
Vancouver
1.
Mets K, Depuydt F, Develder C. Two-stage load pattern clustering using fast wavelet transformation. IEEE TRANSACTIONS ON SMART GRID. 2016;7(5):2250–9.
IEEE
[1]
K. Mets, F. Depuydt, and C. Develder, “Two-stage load pattern clustering using fast wavelet transformation,” IEEE TRANSACTIONS ON SMART GRID, vol. 7, no. 5, pp. 2250–2259, 2016.
@article{8508469,
  abstract     = {{Smart grids collect large volumes of smart meter data in the form of time series, or so-called load patterns. We outline the applications that benefit from analyzing this data (ranging from customer segmentation to operational system planning), and propose two-stage load pattern clustering. The first stage is performed per individual user and identifies the various typical daily power usage patterns (s) he exhibits. The second stage takes those typical user patterns as input to group users that are similar. To improve scalability, we use fast wavelet transformation (FWT) of the time series data, which reduces the dimensionality of the feature space where the clustering algorithm operates (i.e., from N data points in the time domain to log N). Another qualitative benefit of FWT is that patterns that are identical in shape, but just differ in a (typically small) time shift still end up in the same cluster. Furthermore, we use g-means instead of k-means as the clustering algorithm. Our comprehensive set of experiments analyzes the impact of using FWT versus time-domain features, and g-versus k-means, to conclude that in terms of cluster quality metrics our system is comparable to state-of-the-art methods, while being more scalable (because of the dimensionality reduction).}},
  author       = {{Mets, Kevin and Depuydt, Frederick and Develder, Chris}},
  issn         = {{1949-3053}},
  journal      = {{IEEE TRANSACTIONS ON SMART GRID}},
  keywords     = {{IBCN,ELECTRICITY CUSTOMERS,CLASSIFICATION,PROFILES,IDENTIFICATION,SEGMENTATION,CONSUMERS,TIME}},
  language     = {{eng}},
  number       = {{5}},
  pages        = {{2250--2259}},
  title        = {{Two-stage load pattern clustering using fast wavelet transformation}},
  url          = {{http://doi.org/10.1109/TSG.2015.2446935}},
  volume       = {{7}},
  year         = {{2016}},
}

Altmetric
View in Altmetric
Web of Science
Times cited: