A Supervised Learning Approach to Appliance Classification Based on Power Consumption Traces Analysis

Abstract

Electric appliances are everyday major power consumers. Management and control of these appliances can only be possible with appliance classification infrastructures. An appliance classification smart meter, with a provision for remote control, is developed based on time-dependent power features drawn by an appliance, from power-up to its steady state. The kNN classifier is highly accurate at 95.55% in classifying the appliance class.