A Data Driven Approach for Anomaly Detection in Renewable Energy Integrated Power Systems

Abstract

As Distributed Energy Resources (DER) continue to proliferate in power systems, ensuring their reliable and efficient operation becomes paramount. Anomaly detection is a crucial task to maintain the integrity of DER integrated systems and prevent operational disruptions. In this case study, a datadriven approach based Empirical Mode Decomposition (EMD) for anomaly detection in DER integrated systems is presented. Leveraging feature engineering techniques, we analyze the behavior and performance of DER components, including renewable energy sources, energy storage systems, and grid-interconnected devices. This approach utilizes experimental data and real-time measurements to establish normal operating conditions and identify deviations indicative of anomalies, which may encompass equipment malfunctions, load variations, or grid disturbances. Through this case study, the practical implementation is showcased using the proposed method on a real-world DER integrated system, highlighting its ability to enhance system reliability, reduce downtime, and optimize energy management. This case study underscores the efficacy of EMD based anomaly detection in ensuring the reliability of DER integrated systems as they become an integral part of the modern energy landscape. © 2024 IEEE.