Classification of AMI residential load profiles in the presence of missing data

Research output: Contribution to journalArticlepeer-review

3 Citations (Scopus)
98 Downloads (Pure)


Domestic energy usage patterns can be reduced to a series of classifications for power system analysis or operational purposes, generalizing household behavior into particular load profiles without noise induced variability. However, with AMI data transmissions over wireless networks becoming more commonplace data losses can inhibit classification negating the benefits to the operation of the power system as a whole. Here, an approach allowing incomplete load profiles to be classified while maintaining less than a 10% classification error with up to 20% of the data missing is presented.
Original languageEnglish
Pages (from-to)1944 - 1945
Number of pages2
JournalIEEE Transactions on Smart Grid
Issue number4
Early online date28 Apr 2016
Publication statusPublished - 31 Jul 2016


  • load modeling
  • power systems
  • advanced metering infrastructure

Cite this