Stacking Ensemble Method to Predict the Pool Boiling Heat Transfer of Nanomaterial-Coated Surface

Abstract The boiling heat transfer coefficient is important information for designing thermal devices for effective thermal management. It is affected by several factors like surface roughness and wettability of the surface. So, it is necessary to create a model for the accurate prediction. This article aims to use the stacking ensemble method to predict the boiling heat transfer coefficient (BHTC). To improve the performance of the prediction of the stacking model, AdaBoost regression and Random Forest regression are chosen as the base learner, and meta estimator linear regression is selected. Datasets are generated from a pool boiling experiment of carbon nanotube and graphene oxide (CNT + GO)-coated surface. Results have depicted that the stacking method outperformed individual models. It is found that the accuracy of the stacking ensemble model is 99.1% efficient with mean absolute error (MAE), mean square error (MSE), and root mean square error (RMSE) values of 0.016, 0.0004, and 0.021, respectively.