Model-Free Predictive Control Strategy for Photovoltaic Grid-Forming Inverters with MPPT and Power-Dispatch Capabilities

Abstract

As the penetration of photovoltaic (PV) sources into grids increases, transitioning PV grid-connected systems from grid-following (GFL) control to grid-forming (GFM) control is crucial for enhancing active grid support. Owing to the randomness and fluctuation of PV power generation, PV GFM inverters face challenges related to the power mismatch between their AC and DC sides. If the maximum output power of the PV array fails to meet the GFM demand, this insufficiency poses critical challenges for stable GFM operation. A model-free predictive control (MFPC) strategy is proposed for PV GFM inverters with maximum-power point tracking (MPPT) and power-dispatch capabilities. The operating mode of the grid-connected PV system is determined by comparing the maximum PV power with the power required on the AC side. Subsequently, a compact-format dynamic linearization (CFDL) model of the boost converter is established. By adjusting the weight factor of the cost function, the PV current and DC voltage are tracked, enabling adaptive switching between the MPPT mode and power-dispatch mode (PDM). Model-free adaptive control (MFAC) is used for the voltage-current inner loop of the GFM inverter to further enhance the dynamic performance and robustness. Simulation and experimental results indicate that the proposed method enables adaptive switching between the MPPT mode and PDM as the irradiation and AC-side dispatchable power changes. Additionally, the proposed method provides better dynamic performance in the switching process and better robustness when the filtering parameters are mismatched.