Event analysis of distribution system voltage data

Desiree Phillips, Thomas J Overbye

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


Discussion of the future of the Smart Grid brings with it issues regarding the current level of distribution system visibility. Phasor Measurement Units (PMU) have been proliferating throughout the electric power grid in order to support future distribution system analysis. One type of PMU is the Frequency Disturbance Recorder (FDR); it records voltage magnitude, frequency, and phase angle at the 120 V outlet level. Previous work has introduced ways in which distribution system events can be read from this voltage data, but this paper focuses on defining similarity measures for comparing detected events. Discrete Signal Processing (DSP) techniques such as correlation and convolution can be used to determine how similar events are, even in the presence of a time delay. Euclidean distance and mean-squared error are used as similarity measurements to help confirm the correctness of the time-delay estimation. Beyond detecting positively and negatively correlated events, the similarity measures support the correlation results by determining that some events aren't correlated at all.

Original languageEnglish (US)
Title of host publication2015 IEEE Power and Energy Conference at Illinois, PECI 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781479979493
StatePublished - Mar 20 2015
Event2015 IEEE Power and Energy Conference at Illinois, PECI 2015 - Champaign, United States
Duration: Feb 20 2015Feb 21 2015

Publication series

Name2015 IEEE Power and Energy Conference at Illinois, PECI 2015


Other2015 IEEE Power and Energy Conference at Illinois, PECI 2015
Country/TerritoryUnited States


  • Correlation
  • Distribution Systems
  • Electric Power
  • Frequency Disturbance Recorder
  • Signal Processing
  • Voltage

ASJC Scopus subject areas

  • Energy Engineering and Power Technology
  • Fuel Technology


Dive into the research topics of 'Event analysis of distribution system voltage data'. Together they form a unique fingerprint.

Cite this