Thermal efficiency of Parabolic Trough Collector Working with Al₂O₃ /water Nanofluid: Experimental and Machine Learning Analysis

The thermal efficiency of a parabolic trough collector working with $\mathrm{Al}_{2} \mathrm{O}_{3} /$ water nanofluids at different mass flow rates, and volume concentrations have been estimated experimentally. The experiments were performed under the mass flow rate range from $0.5 \mathrm{lit} / \mathrm{min}$ to $2 \mathrm{lit} / \mathrm{min}$, and the volume concentrations range from $0.05 \%$ to $0.5 \%$, respectively, and also the experiments were conducted during the daytime from 10.00 hours to 16:00 hours. All the measurements were performed based on the ASHRAE 93 standards. Results indicated that the collector efficiency is increased by $10.78 \%, 11.33 \%$, and $12.28 \%$ at a flow rates of $0.5 \mathrm{lit} / \mathrm{min}, 1 \mathrm{lit} / \mathrm{min}$, and $2 \mathrm{lit} / \mathrm{min}$ and at a volume concentration of $0.5 \%$ compared to water data. The obtained thermal efficiency is used for Extreme Gradient Boosting (XGBoost) machine learning algorithm for predictions. The regression coefficients of all data for fluid flow rate is $0.0083 \mathrm{~kg} / \mathrm{s}$ is 0.99782, for $0.0167 \mathrm{~kg} / \mathrm{s}$ is 0.99626, and for $0.033 \mathrm{~kg} / \mathrm{s}$ is 0.99575, respectively. Based on the efficiency data, multilinear regression analysis is for developing the equations.