مشخصات پژوهش

The increasing complexity of modern DC microgrids, driven by the integration of renewable energy and storage units, makes their analysis and simulation both challenging and time-consuming. Dynamic equivalencing has emerged as an effective solution to reduce model complexity, accelerate simulations and improve understanding of system dynamics. This paper presents a circuit-based equivalent model for a grid-connected DC microgrid using a grey-box approach that bridges physical modelling and data-driven identification. The proposed model is operating-point independent and capable of accurately reproducing both small- and large-signal dynamics. It preserves the original circuit structure, while its parameters are identified from measurement data at the point of common coupling and MATLAB/Simulink simulations. Parameter estimation is carried out under diverse scenarios — including load variations, solar irradiance fluctuations and short-circuit faults — using a genetic algorithm for optimal identification. The studied microgrid comprises multiple photovoltaic units, battery storage systems and resistive loads; in the equivalent model, one representative unit of each type is employed with a scaled capacity. The proposed model is also benchmarked against a conventional black-box model. Validation results show that the proposed grey-box equivalent can faithfully reproduce the dynamic behaviour of the detailed microgrid, achieving an R2 index above 90% across all scenarios, demonstrating its suitability for future control and operational studies. This makes the proposed equivalent particularly useful for controller design, stability assessment and real-time simulation of DC microgrids.