Sim-to-Real Global Maximum Power Point Tracking With Domain Randomization and Adaptation for Photovoltaic Systems
Simulations of photovoltaic (PV) systems help understand the nonlinear power-voltage characteristics in real-world atmospheric conditions. However, the gaps between simulation and real-world domain are usually significant due to the modeling errors. Therefore, we propose a simulation-to-reality (sim-to-real) global maximum power point tracking (GMPPT) method with domain randomization and adaptation for PV systems. The randomized PV model is adapted to the real-world data domain to estimate the operating point at which the maximum power is drawn, and the sim-to-real gap is bridged via a vortex search algorithm. With the consolidated estimation results, the proposed method can adapt to the dynamics of the real-world environment and accelerate the GMPPT process. Experimental results show that the proposed method can reduce sim-to-real discrepancies and enhance GMPPT efficiency in comparison to the existing methods.
