Optimization assessment and performance analysis of an ingenious hybrid parabolic trough collector: A machine learning approach

Abstract

Parabolic Trough Collector (PTC) is a solar unit that can convert solar radiation into thermal energy at moderate efficiency. Researchers attempted to increase the output of the PTC units by implementing multiple methods including using helical inserts, transparent aerogel, and reflective mirrors. Nevertheless, the output of this system still is not high enough. The proficiency of the PTC unit can improve by integrating it with Thermoelectric Generator (TEG) and Proton Exchange Membrane (PEM) electrolysis cell, which is studied in the current paper for the first time. In the present study, the efficiency in both PTC-TEG and PTC layouts are poked to demonstrate the consequence of the TEG on the denoted unit efficiency. Besides, the sensitivity study is carried out to assess the influence of operating parameters. Also, extensive data collection is generated and employed to train the designed machine learning models to anticipate the performance of the units in countless operating conditions. Based on the results, the highest accuracy in predicting the performance of the units belongs to the adaptive network-based fuzzy inference system. Also, the results show that the highest entropy generation rate for the PTC and PTC-TEG-PEM units is approximately 11.0 W/m.K and 16.9 W/m.K, respectively. In addition, the highest hydrogen production rate of the PTC-TEG-PEM unit is around 21 mol/h; in fact, the electrical input of the PEM electrolyzer is maintained by the TEG module, which its electrical efficiency calculated to be 3%.