Intelligent home energy management using Internet of Things platform based on NILM technique

Abstract

Due to the continuous increase in the global energy demand, it is essential to find solutions to improve energy efficiency. Non-intrusive load monitoring (NILM) is one of the most prominent energy management techniques, and it allows consumers and companies to manage their energy consumption efficiently. The energy footprint of residential buildings is monitored, and the total electrical consumption is disaggregated into appliance related signals through NILM. The NILM applications are working based on the “energy awareness” concept where the consumer receives some itemized information from the application to reduce its consumption and thus improves energy efficiency. This paper presents an Internet of Things (IoT) based criterion by using NILM to transform modern buildings and homes into energy-efficient and smart ones. The Factorial Hidden Markov Model (FHMM) is used as a NILM technique to disaggregate all appliances’ total power consumption into individual appliance load consumption by giving the power as an input. The FHMM is applied on reference energy disaggregation dataset (REDD) datasets, and a comparison is made between the measured and predicted values of appliances’ power consumption. Instant visualization of live data is then created using the ThingSpeak platform, which sends alerts through Twitter social media application to the consumer to control his consumption. The consumption behavior is assessed according to the time of use to make an energy saving based on the cost of electricity.