Prediction of jet impingement solar thermal air collector thermohydraulic performance using soft computing techniques
Raj Kumar, Nitisha Sharma, Chahat, Gia Huy Ngo, Amit Kumar Yadav, Daeho Lee, Tej Singh
Abstract
Solar thermal air collectors are economically efficient for the purpose of heating air in specific applications. This research examines the efficiency of a solar thermal air collector called a jet impingement solar thermal air collector (JISTAC) that is fitted with discrete multi-arc-shaped ribs (DMASRs) utilizing soft computing techniques. Linear Regression (LR), M5P, Gaussian Process (GP), and Random Forest (RF) are used to predict the Nusselt number (Nu), friction factor (f), and thermohydraulic performance (ηthp) of JISTAC. DMASRs absorber plates with different relative rib height (0.025–0.047), relative discrete width (0.32–1.72), arc angle (35°–65°), relative discrete distance (0.27–0.86), and relative rib pitch (0.58–3.1) were used. The tests yielded 245 data sets, with 173 assigned for training and 72 for soft computing testing. The GP surpasses other models in performance owing to its minimum error and maximum correlation coefficient value. The GP model has a MAE of 0.0347, a RMSE of 0.0564, a RAE of 10.88%, a RRSE of 12.77%, and a CC value of 0.9920. Furthermore, it has a model efficiency of 99%, making it the most prominent among the other models. The results indicate that the GP model is quite precise in predicting the ηthp of JISTAC.