ComputingElectricityConsumptionProlesfromHousehold.pdfVIP

Computing Electricity Consumption Profiles from Household Smart Meter Data ⇤ Omid Ardakanian Negar Koochakzadeh Rayman Preet Singh University of Waterloo Oracle University of Waterloo Waterloo, Ontario, Canada Vancouver, BC, Canada Waterloo, Ontario, Canada oardakan@uwaterloo.ca negar@ rmmathar@uwaterloo.ca Lukasz Golab S. Keshav University of Waterloo University of Waterloo Waterloo, Ontario, Canada Waterloo, Ontario, Canada lgolab@uwaterloo.ca keshav@uwaterloo.ca ABSTRACT during peak times. In this paper, we investigate a critical problem in smart However, exploiting smart metering systems to their fullest meter data mining: computing electricity consumption pro- also requires mining the vast amounts of collected consump- files. We present a simple, interpretable and practical pro- tion data to obtain insights into grid operations and con- filing framework for residential consumers, which accounts sumer behaviour [2, 4, 7, 12, 16, 21]. for variations in electricity consumption at di↵erent times In this paper, we address the problem of computing elec- of day and at di↵erent external temperatures. Our approach trici