Pattern Analysis of Residential Electricity Consumption Using Circuit Level Monitoring Data

Abstract

Energy management systems (EMS) on demand sides are expected as a method to enhance the capability of supply and demand balancing of a power system under the lack of sufficient electricity supply capacity after the Great East Japan Earthquake and increasing renewable energy generation such as Photovoltaics (PV). EMS is supposed to determine the optimal daily operation of energy appliances including energy storage factors economically and efficiently to satisfy user's desire taking account of forecasted energy demand and supply given by renewables such as PV. Energy demand forecast, being one of important factors to realize EMS, it is very difficult because building energy consumption structure is not elucidated sufficiently due to the variation of appliances and life styles. We need the information which kind of electricity demand has a likelihood of activation (shifted or reduced) for the forecast in EMS. Grasping energy consumption structure in a building contributes to not only demand forecast but also extraction of potential energy saving.
In this paper, we propose pattern recognition method of circuit currents including periodic or spike wave obtained by circuit level monitoring.