Neural network combined with nature-inspired algorithms to estimate overall heat transfer coefficient of a ribbed triple-tube heat exchanger operating with a hybrid nanofluid

Abstract

Four nature-inspired optimizers are combined with a multilayer perceptron neural network for reaching an optimal structure aimed at predicting the overall heat transfer coefficient of a ribbed triple-tube heat exchanger in terms of the rib pitch, rib height, and nanoparticle concentration. The heat exchanger works with a hybrid nanofluid having graphene nanoplatelet/Pt nanocomposite powder. The applied algorithms incorporate Artificial Bee Colony (ABC), Ant Colony Optimization (ACO), Ant Lion Optimizer (ALO), and Biogeography-Based Optimization (BBO). The required data are provided via computational solutions. The BBO algorithm is determined as the best method to predict the output because of its higher accuracy. The highest performance of the BBO is obtained from population of 450. The overall heat transfer coefficient is estimated with root mean square errors of 0.030 and 0.025 for the training data and testing data, respectively. Furthermore, the ACO algorithm shows the lowest computational time among all the methods.