A New Intelligent System Architecture for Energy Saving in Smart Homes

Abstract

Technologies to support the development of smarter energy systems and increase the opportunities for home energy management have increased in recent years. Through the addition of sensing, communication and drive components, devices and home appliances become increasingly smart so that they can communicate with each other, transmit data to end users and facilitate remote operation and automation, for example during periods of peak demand. This has the potential to provide energy-related benefits to end-users and network operators. One of the key benefits of intelligent techniques is the potential to support power reductions and demand side management. Within this context, this work aims to propose a system architecture for recommend suggestions to smart homes inhabitants (according to different user profiles), aiming cost reduction and energy efficiency. The work also includes the implementation of a RESTFul mining web service, which is part of the proposed architecture. The web service is used to release data mining, classifying data over the proposed recommender system. Through the data model transmission known as JSON (JavaScript Object Notation), the web service receives data and automatically convert to JSON Instances structures, processing the same with the data mining algorithms.