Design and Implementation of an IoT-Oriented Energy Management System Based on Non-Intrusive and Self-Organizing Neuro-Fuzzy Classification as an Electrical Energy Audit in Smart Homes

Abstract

Smart cities are built to help people address issues like air pollution, traffic optimization, and energy efficiency. Electrical energy efficiency has become a central research issue in the energy field. Smart houses and buildings, which lower electricity costs, form an integral part of a smart city in a smart grid. This article presents an Internet of Things (IoT)-oriented smart Home Energy Management System (HEMS) that identifies electrical home appliances based on a novel hybrid Unsupervised Automatic Clustering-Integrated Neural-Fuzzy Classification (UAC-NFC) model. The smart HEMS designed and implemented in this article is composed of (1) a set of IoT-empowered smart e-meters, called smart sockets, installed as a benchmark in a realistic domestic environment with uncertainties and deployed against non-intrusive load monitoring; (2) a central Advanced Reduced Instruction Set Computing machine-based home gateway configured with a ZigBee wireless communication network; and (3) a cloud-centered analytical platform constructed to the hybrid UAC-NFC model for Demand-Side Management (DSM)/home energy management as a load classification task. The novel hybrid UAC-NFC model proposed in DSM and presented in this article is used to overcome the difficulties in distinguishing electrical appliances operated under similar electrical features and classified as unsupervised and self-organized. The smart HEMS developed with the proposed novel hybrid UAC-NFC model for DSM was able to identify electrical household appliances with an acceptable average and generalized classification rate of 95.73%.

Keywords

• IoT
• load disaggregation
• neuro-fuzzy classification
• non-intrusive load monitoring
• unsupervised learning
• smart homes