Modeling and Experimental Validation of a Solar Panel for Digital Twin Applications

Abstract

As the global demand for renewable energy grows, effective management of photovoltaic (PV) systems is essential as the world's need for renewable energy increases. This paper presents the modeling and experimental validation of a 60 W monocrystalline photovoltaic module, with a focus on digital twin applications for photovoltaic system management, particularly energy generation forecasting and fault detection. The proposed model is based on the Single Diode Model (SDM), an electrical modeling approach widely adopted due to its balanced trade-off between simplicity and accuracy, and experimental validation of a 60 W monocrystalline PV module. A hybrid optimization technique that combines Powell's method and Particle Swarm Optimization (PSO) was created to estimate model parameters. Compared to standard PSO methods, which usually require 1,000 iterations, this hybrid approach achieved high accuracy in just 100 iterations. The modeling was implemented in MATLAB/Simulink and experimentally validated using data collected under two solar irradiance conditions (1000 W/m² and 500 W/m²), at 25°C and air mass AM1.5, using a flash-type solar simulator. The current-voltage (I-V) and power-voltage (P-V) curves showed remarkable precision, with R2 values exceeding 0.998 and with MAE values of 0.2932W. Based on these results, the developed model proves to be highly suitable for digital twin-based applications, especially in scenarios that require high reliability, such as energy forecasting and real-time fault diagnosis. Furthermore, the developed experimental dataset is made publicly to support researchers who do not have access to specialized laboratory infrastructure.