Optimized model predictive control with gray-box modeling for PV-based DC/DC converters

This study presents an advanced model predictive control (MPC) framework integrated with a gray-box modeling approach for voltage stabilization in photovoltaic (PV) power systems utilizing a DC/DC boost converter. As a continuation of our previous research, where a gray-box model was developed to accurately represent the boost converter dynamics, this work extends the model's application by designing and implementing a predictive controller. The gray-box model, which combines theoretical equations with empirical system identification, serves as the foundation for the proposed MPC, ensuring precise control action under varying environmental and load conditions. Simulation results indicate that the MPC significantly outperforms conventional proportional-integral-derivative controllers, reducing output voltage ripple to near zero and achieving a settling time of 0.0002 s. Experimental validation confirms the controller’s robustness, maintaining stable voltage regulation under dynamic conditions with minimal transient effects. The findings demonstrate the effectiveness of gray-box-based MPC in enhancing the reliability and efficiency of PV power systems, paving the way for future advancements in intelligent control strategies for renewable energy applications. • Gray-box modeling integrated with MPC for PV boost converters. • Gray-box improves accuracy by merging theory and experiments. • MPC minimizes ripple and achieves 0.0002 s settling time. • Experiments confirm robust control under solar/load changes. • Real-time hardware tests prove practical feasibility.