Event-Triggered Collaborative Control Models for Microgrid Systems

The unique ability to combine distributed renewable energy sources and enhance system reliability, resilience, and system flexibility has made microgrids a key component of modern power system. But the existing time-triggered control approaches involve periodic communication between the distributed controllers irrespective of the system status which leads to unnecessary communication overhead, high energy consumption, and low scalability. To overcome these limitations, this paper proposes an Event-Triggered Collaborative Control Model (ET-CCM) that combines Random Forest (RF)-based event detection with Federated Learning (FL)-based collaborative learning. The RF classifier identifies critical operating conditions, such as voltage and frequency deviations, and allows for control actions to take place only when needed, substantially minimizing unnecessary communications.Simultaneously, FL allows multiple microgrids to collaboratively train a global model by exchanging model parameters instead of raw operational data, ensuring privacypreserving distributed learning. The proposed ET-CCM is implemented and evaluated in the MATLAB environment under different operating scenarios involving renewable energy intermittency and load variations. Simulation results demonstrate that the proposed model reduces communication load by approximately 30% while improving voltage regulation accuracy by about 20% compared with conventional time-triggered control. Furthermore, the Voltage Stability Index (VSI) improves from 0.45 to 0.38 (approximately 18% improvement), while the Frequency Regulation Error (FRE) decreases from 3.5% to 2.5% (approximately 15% improvement). These results confirm that the proposed ET-CCM provides improved communication efficiency, enhanced voltage and frequency regulation, better system stability, and privacy-preserving collaborative control for interconnected microgrid systems.