Experimentally validated thermal modeling for temperature prediction of photovoltaic modules under variable environmental conditions

Abstract

In this work, a detailed analysis and thermal modeling for temperature prediction of a stand-alone photovoltaic module is performed. The study aims to present precise estimation of module temperature, since it is an important parameter for power output calculation. Hence, the required data were collected via experiments. Accounting for all heat transfer mechanisms, and following model validation, a proposed algorithm was implemented to investigate heat transfer from the module to its surrounding and predict different layers’ temperature. Results indicate that accurate energy distribution and temperature prediction was achieved by the adopted thermal model, only about $16\%$ of the received energy is converted to electrical power while the rest is released by heat. Moreover, the proposed simulation algorithm provided one of the best results in comparison to literature models, achieving an $R^2$ of $0.963$ and a $MAE$ of $1.883$, which is very close to the best overall model by King at $R^2=0.973$ and $MAE=1.663$. Additionally, two new models for module temperature prediction were proposed. After testing on new data, the explicit model provided a reasonable first approximation attaining an adjusted $R^2$ of $0.97$ and a $MSE$ of $3.505$, and an accurate implicit model, achieving a $MSE$ of only $1.268$.