Thermal Mathematical Modeling of Photovoltaic Inverters and Experimental Validation

• Published in: Brazilian Power Electronics Conference (Congresso Brasileiro de Eletrônica de Potência) pp. 1 - 7
• Main Authors: de Lima, Geyciane Pinheiro, Schmitt Prym, Guilherme Cerbatto, de Melo, Karen Barbosa, Moreira, Hugo Soeiro, de Souza Silva, Joao Lucas, Filho, Ernesto Ruppert, Villalva, Marcelo Gradella
• Format: Conference Proceeding
• Language: English
• Published: IEEE 07.11.2021
• Subjects: Analytical models; Electric variables measurement; Inverters; Linear regression; photovoltaic; Photovoltaic systems; prediction of temperature; Predictive models; Temperature dependence; Thermal mathematical modelling
• ISSN: 2643-9778
• DOI: 10.1109/COBEP53665.2021.9684058

The inverter is one of the most failure-prone components of the photovoltaic (PV) system. Failure can happen under the influence of variables such as temperature, making it a challenge to study temperature behaviour under different conditions. It is possible to extend the useful life and increase the reliability of the inverter by investigating the thermal influence on electrical parameters and by evaluating the temperature under different operating conditions. Mathematical thermal modeling is an important tool in the analysis of the operating behavior of the solar inverter. Papers on this subject are rare in the literature. This work analyzes an inverter thermal modeling method and validates it through laboratory experiments. This work also makes a comparison with another method found in the literature, based on linear regression. As a result, the mean absolute percentage error of the analyzed modelling method was less than 2 % in comparison with data obtained experimentally. An important conclusion of this work is that a good thermal modeling requires the experimental determination of the thermal coefficients that integrate the mathematical model.