Maximizing thermal and electrical efficiency with thermoelectric generators and hybrid photovoltaic converters: Numerical, economic, and machine learning analysis
Haitham, Osman
Loke Kok, Foong
Binh Nguyen, Le
Spalević, Velibor
Dudić, Branislav
Škatarić, Goran
Abstract: In this paper, we introduce an innovative thermoelectric, photovoltaic hybrid system and investigate its performance under various radiation intensities and heat transfer coefficients outside the cavity. Our findings reveal that the proposed system yields twice the power output compared to a traditional plate thermoelectric, photovoltaic hybrid system. Through economic analysis, we project a 45 % reduction in energy cost with this novel structure compared to a full hybrid system. Notably, positioning the hybrid system at the bottom of the cavity, where maximum radiation occurs, is deemed optimal. Our heat transfer analysis demonstrates a significant increase in power generation due to convection outside the cavity, with approximately 9 % of incoming radiation reflected and a further 59 % reflected without the cavity. Utilizing artificial neural networks, we predict thermal and electrical power generation, achieving a Mean Absolute Error (MAE) below 3 % and an R-squared value exceeding 0.98. Additionally, our model's predictions closely match experimental results, validating its accuracy and practical utility. This comprehensive study advances the field by offering a novel hybrid system design that outperforms existing solutions while providing insights into optimizing placement and enhancing power generation through sophisticated modeling techniques.
engleski
2024
Ovo delo je licencirano pod uslovima licence
Creative Commons CC BY 4.0 - Creative Commons Autorstvo 4.0 International License.
http://creativecommons.org/licenses/by/4.0/legalcode
Keywords: Hybrid systemPhotovoltaic-thermoelectricSolar absorptionPower generationArtificial neural network
o:1600 | Radovi profesora i saradnika Fakulteta za ekonomiju i inženjerski menadžment |