Big Data Architecture for Efficient Energy Management in Multi Microgrid Scenarios

The traditional electric power distribution network is evolving to support scalable and connected services in Smart Cities (SC). The emergence of Smart Grid (SG) architecture connects Smart Meters (SM) to the Electric Power Company (EPC) using various communication technologies like Power Line Communication (PLC) and wireless networks such as 5G, LoRa, and others. However, the vast amount of data generated by this infrastructure can lead to network overload and storage and processing challenges. To address this, this paper proposes implementing a Big Data Analytics framework under the Hybrid Demand Side Management as a Service (HyDSMaaS) architecture in the Multi Microgrid scenario. This approach divides the SG into Microgrids, creating smaller groups of nodes and transforming the centralized infrastructure into a scalable and decentralized one. Additionally, a monitoring system using Grafana was implemented to manage server usage flow and monitor access and services. The framework achieved high speed compared to traditional models and demonstrated accuracy of nearly 99.9% in consumption prediction and power quality classification algorithms when operating with up to 1000 microgrids in under 1000 seconds.