Ecological and secure electricity microgrids - monitoring and forecasting challenges
Wiktor Wojciechowski, Adam Niewiadomski, Yuriy Bilan
Abstract
Reliable, low-carbon power systems depend on high-granularity, short-horizon forecasting, especially within islandable microgrids where consumers and prosumers shape real-time balance. Ultra-short-term (15-minute) forecasts support storage scheduling, demand response, and secure operation amid renewable intermittency and growing cyber risk. Forecasting approaches are classified into black-box, gray-box, and white-box methods, highlighting trade-offs in accuracy, explainability, and deployability. Operational use-cases are aligned with forecasting time scales, and current literature shows notable gaps: limited 15-minute multi-step studies, inconsistent evaluation protocols, insufficient attention to explainability, and restricted access to representative datasets. Design principles are outlined for deployable forecasting and control: standardised metrics and horizons, privacy-preserving data pipelines, explainability-first modelling, and transferable domain-specific hyperparameters. Addressing these gaps can increase renewable penetration, lower imbalance costs, strengthen cybersecurity compliance, and enhance resilience, delivering cleaner energy at reduced cost with higher quality of service. Researchers who want to support effective energy management technologies with their research should be aware of the current challenges.