Modeling and optimization of back temperature models of mono-crystalline silicon modules with special focus on the effect of meteorological and geographical parameters on PV performance

In this study, experimental technique was used to obtain datasets applied for developing 80 models, optimizing 7 models and computing back temperature at various tilt angles and orientations in Lagos, Nigeria and Africa. Statistical evaluation and ranking tools such as MBE, MPE, RMSE, RRMSE, R2 and GPI results revealed that model 3 outperformed excellently for module 3, model 31 for module 4, model 47 for module 5, model 57 for module 6 and model 70 for the generalised module category and equally the best for the 80 established models. Similar statistical analysis was performed on the selected seven models obtained from literature so as to test their performances with best established model in each category of development. The result revealed that the established models in this study outperformed the seven unoptimised models from literature. The authors thereafter established a single hybrid-parameter based optimization based model so as to enhance the performance capacity of the seven models. The result equally revealed that the established best performing models in each category significantly outperformed these seven models. The newly established monthly and yearly module temperature coefficient models performed excellently in the 36 towns in Nigeria and 54 cities in Africa when compared ambient temperature.