Appliance Classification Method Based On K-Nearest Neighbors for Home Energy Management System

Saving energy is one of the biggest concerns nowadays in order to deal with lacking fossil fuel and climate change. One possible approach is using the Home Energy Management System (HEMS) to monitor and manage the electrical usage at residential user. One useful feature of HEMS is the capability of detecting and identifying electrical appliances in order to assign these appliances to join many saving energy programs such as Time-Of-Use or Demand Response. In this paper, we propose an appliance classification approach based on K-Nearest Neighbors (KNN) technique. KNN is extensively used classification algorithm owing to its simplicity, ease of implementation and effectiveness. The inputs we used to calculate the Euclidean distance in KNN are the root mean square (RMS) of voltage, current and the active power of appliance. We also implement a HEMS prototype includes many smart plug-in devices and a MQTT server. The appliance classification results show that our KNN classifier can be a suitable approach for appliance classification feature of HEMS.