Dynamic Threshold-Based Event-Triggered Strategy for Robust Fully Distributed Control in Renewable-Powered DC Microgrids

This paper presents a novel Fully Distributed Adaptive Observer-Based Event-Triggered (FDAOET) secondary control strategy for islanded DC microgrids (MGs) that achieves accurate voltage restoration and proportional current sharing while optimizing communication efficiency without compromising system resilience. Unlike conventional static or heuristic event-triggered (ET) schemes, and even existing dynamic ET methods that rely on fixed structures or global parameters, the proposed strategy employs an observer-based, state-aware triggering mechanism. Each distributed generator (DG) transmits data only when the deviation between its internal observer estimate and local measurement exceeds an adaptively evolving threshold governed by an ordinary differential equation (ODE). This formulation enables robust, context-aware communication scheduling that maintains estimation accuracy under noise, parameter uncertainties, and network variability. Additionally, the proposed method explicitly incorporates bounded communication and actuation delays into both the triggering and control design, ensuring reliable operation under realistic conditions. The FDAOET strategy is fully distributed and topology-independent, requiring no global information such as Laplacian matrices, thereby supporting plug-and-play scalability. Zeno behavior is rigorously avoided, and system stability is proven using Lyapunov methods. Simulation results in MATLAB/SimPowerSystems demonstrate superior performance in terms of convergence speed, voltage accuracy, current sharing, and significant reduction in communication events compared to static, dynamic, and delay-unaware ET methods.