A Real-Time Cyber-Physical Emulation Platform for Scalable Anomaly Detection and Monitoring in Smart Microgrids

This paper presents a real-time smart microgrid (SMG) emulation and anomaly detection platform for evaluating intelligent monitoring methods under realistic operating conditions. The distributed cyber-physical nature of SMGs makes them vulnerable to operational and cyber anomalies, including voltage and frequency deviations, abnormal load changes, renewable generation fluctuations, battery faults, sensor failures, controller malfunctions, communication disruptions, data corruption, and coordinated cyberattacks. However, existing anomaly detection methods are commonly validated using offline datasets, simulations, or isolated microgrids rather than real-time interconnected SMG environments. They also face limitations in scalability, continuous monitoring, anomaly localization, detection of simultaneous multidimensional anomalies, adaptation to non-stationary operating conditions, and dependence on labeled datasets. To address these limitations, this paper introduces an integrated platform comprising the Common Open Research Emulator, the Zero Network Overhead communication protocol, InfluxDB, Grafana, and Python-based analytical modules. The platform supports scalable and low-latency data exchange, storage, visualization, and analysis across multiple interconnected SMGs. Its emulation environment enables anomaly detection methods to be evaluated under realistic and continuously changing operating conditions without requiring a fully deployed physical microgrid infrastructure. Furthermore, a multidimensional discriminant analysis-based method (MDAM) is incorporated to provide robust anomaly detection by jointly analyzing changes in the mean and variance of heterogeneous operational measurements. Experimental results obtained under multiple real-time operating scenarios demonstrate accurate detection and localization of unidimensional and multidimensional anomalies, continuous system visualization, and scalable performance. The proposed platform bridges the gap between theoretical anomaly detection research and practical SMG deployment by providing a flexible, reproducible, hardware-independent environment for developing and validating intelligent monitoring technologies.