Model Predictive Control of Independent Photovoltaic Systems Under Complex Weather Conditions Based on Energy Functions

Abstract

This paper proposes a control strategy for a standalone photovoltaic single‐phase split‐source inverter based on energy function control. Unlike conventional predictive control, the proposed method utilizes the inherent redundancy in energy function coefficients, thereby avoiding the complex tuning of weighting factors. The system consists of a photovoltaic array, a simplified split‐source inverter, an energy storage battery, and a bidirectional DC–DC converter. Maximum power point tracking is realized through a perturbation and observation method incorporating PI control, while the split‐source inverter provides both MPPT and a sinusoidal output voltage. A PI controller within the bidirectional DC–DC converter maintains DC‐side voltage stability. Simulation results confirm that the system achieves maximum photovoltaic power output and stable load voltage under varying operating conditions, while maintaining low total harmonic distortion. Experimental results further confirm both the method's effectiveness and overall system performance.