Estimation models of photovoltaic module operating temperature under various climatic conditions

Abstract

Accurate control of the operating temperature is essential for achieving optimal efficiency in the use of solar panels. The Normal Operating Cell Temperature (NOCT) is a widely used method for estimating module temperature, but it is not always accurate for all conditions. In this paper, the authors propose four different models for estimating module temperature using various techniques including Neural Networks, a linear method proposed by Ross, and a Fitting method. The objective is to find more accurate methods than the NOCT model. The results show that all four models provide good agreement between measured and calculated module temperatures, with R² > 0.91 and RMSE < 3.74 for clear days, and considering the resources of time, cost, and expertise necessary for simulating or evaluating the proposed model, the two models based on RN neural networks offers a more cost-effective solution, by utilizing straightforward mathematical skills, the RN model demonstrates applicability in any environment, making it a preferable choice over the NOCT model for predicting polycrystalline photovoltaic module temperature under different climatic conditions. The proposed models can be used to estimate PV module temperature with good precision under different climatic conditions.