Adaptive MPC-Based Grid-Forming Control for Hybrid Renewable Storage Systems

Abstract

The paradigm shift toward 100% renewable energy systems has precipitated a critical decline in rotational inertia, rendering modern power grids highly susceptible to frequency instability and high Rates of Change of Frequency (RoCoF). Grid-forming (GFM) inverters utilizing Virtual Synchronous Generator (VSG) control offer a promising solution; however, existing methodologies predominantly employ fixed virtual inertia and damping coefficients, which fail to exploit the dynamic capabilities of hybrid renewable energy systems (HRES). This paper addresses this gap by proposing a Linear Parameter-Varying Model Predictive Control (LPV-MPC) framework for adaptive VSG control in hybrid solar photovoltaic (PV), wind, and battery energy storage systems (BESS). The proposed methodology dynamically optimizes virtual inertia and damping in real time by formulating the control objective as a constrained Quadratic Program (QP). The optimization explicitly accounts for the stochastic generation limits of the PV-wind system, inverter capacity constraints, and the real-time State-of-Charge (SoC) of the BESS. By embedding the nonlinear VSG dynamics into an LPV state-space model, the proposed MPC guarantees optimal transient performance while preventing BESS degradation. Numerical validation on a modified IEEE 39-bus test system demonstrates that the proposed LPV-MPC adaptive VSG mitigates maximum frequency deviation by 38% and RoCoF by 42% compared to conventional fixed-parameter VSG, ensuring robust operation under severe low-inertia conditions and asymmetrical faults.