Energy Disaggregation Using Energy Demand Model and IoT-Based Control

Energy Management Systems involve monitoring of loads, control, and providing recommendations to reduce demand or energy costs. Energy Disaggregation works on monitoring loads nonintrusively, by having a single-smart meter at the entry point to perform the task with machine learning techniques. Training of the machine learning model is an important step and may require historical submetered data of the appliances. In this article, an energy demand model is used to generate the training data and alleviate the need for historical data of the appliance. The model is optimized for the Indian scenario based on the representation of appliances and active occupancy. The other important contribution of this work is the use of Internet of Things (IoT) devices to feed observable states to the disaggregation model to improve efficiency. A selectively enabled factorial hidden Markov model is utilized in which states of IoT control relays are presented to the model. The platform developed, enables both monitoring and control of appliances and provides insight into the overall user energy consumption and its breakdown at the appliance level.