In order to enhance electricity conservation in households, detecting which electric appliances consume high electricity is effective. However, collecting every data of electric appliances needs many devices and it would be high costs. Therefore, data disaggregation from total data to each appliance leads to the reduction of cost and electricity. This paper presents how to disaggregate the consumption of electric appliances from total electricity consumption. In order to disaggregate electricity consumption of electric appliances from total energy consumption, sparse coding has been implemented. However, changes in the use of electric appliances are difficult to express in this method. Therefore, we propose a novel sparse coding method, named, "0-1 sparse coding" to disaggregate which electric appliances were used in the total consumption of electricity. In order to collected sample data for analyzing the methods, we installed smart metering systems in two households. The system collects the data of total electricity consumption and data in the main electric appliances in every 5-minutes. We evaluate two methods, 1) our proposed method, and 2) discriminative sparse coding method. From the results, our proposed method increased the accuracy about 44.8% than the previous method.
|出版ステータス||Published - 2015|
|イベント||7th International Conference on Applied Energy, ICAE 2015 - Abu Dhabi, United Arab Emirates|
継続期間: 2015 3 28 → 2015 3 31
ASJC Scopus subject areas