Model Reference Adaptive Control-Based Virtual Synchronous Generator for Modern Power Systems

The increasing penetration of inverter-based renewable energy sources (RES) has substantially reduced the inertia of modern power systems, posing significant challenges for maintaining frequency stability under disturbances. Virtual Synchronous Generators (VSGs) have been proposed as effective grid-forming solutions; however, conventional VSG control frameworks typically rely on pre-assumed, fixed controller gains and static network conditions. This assumption makes them highly susceptible to dynamic, time-varying grid configurations and fluctuating RES availability, which critically impact system stability margins, power-sharing accuracy, and transient performance, particularly during large disturbances. This paper presents a novel modular VSG control strategy based on Model Reference Adaptive Control (MRAC) to address these limitations. The proposed approach adaptively adjusts control inputs in real time, without the need for explicit estimation or identification of system parameters. A Lyapunov-based adaptive law ensures closed-loop stability and asymptotic tracking of a predefined reference model. Comprehensive time-domain simulations validate the superior frequency regulation, enhanced transient response, and improved robustness of the proposed MRAC-based VSG over the conventional VSG control strategy.