Cloud-Fog-based approach for Smart Grid monitoring

Abstract

Continual, safe, secure, reliable, and efficient delivery of high-quality power is the main goal of the Smart Grid (SG) defined as the energy network integrated with the novel Information and Communication Technologies (ICTs). This paper focuses on the development of communication system models to be used for voltage profile monitoring and power losses estimation in SG. Existing research on communication technologies in SG is briefly overviewed with the main focus on Cloud and Fog computing approaches. With rising numbers of smart devices in SG, the amount of measurement data rapidly increases demanding more efficient communication architectures. Two communication architectures are selected and modeled with the main goal to support real-time monitoring and data analysis in SG. Cloud- and Cloud-Fog-based approaches are proposed and analyzed as communication system models for SG real-time monitoring. Posting immense quantities of data on Cloud commonly represents a serious challenge due to latency, locality, and network congestion. The best approach to overcome Cloud challenges is to perform decentralized and intelligent processing closer to the location where data are being generated, which is known as Fog computing. A MATLAB/Simulink-based model of a well-known IEEE test grid topology is modified to support real-time communication with open source IoT platform ThingSpeak used for Cloud computing. Fog computing layer is conceptualized to be located in the main distribution substation and it is modeled by MATLAB function blocks. The proposed modeling procedure is given step by step to highlight the most important building blocks necessary for creating a combined power and communication system model capable of providing meaningful data related to SG monitoring. The developed model is tested using simulation events common in distribution networks. The simulation results of Cloud and Cloud-Fog approaches for voltage profile monitoring and power losses calculation in SG have confirmed the benefits of the Cloud-Fog-based approach.