A Predictive Approach for Evaluating Thermo-Physical Properties of Nano fluids Using Artificial Intelligence Algorithms

Abstract

Artificial Intelligence (AI) algorithms are increasingly being employed as substitutes for conventional methods or as components within integrated systems. They have demonstrated effectiveness in addressing complex applied problems across various domains, gaining popularity in the present context. AI approaches exhibit the ability to learn from patterns, tolerate faults by handling noisy data, and manage non-linear problems. Once trained, they excel in generalization and fast estimation. This survey presents a comprehensive review of AI algorithms developed for investigating nanofluid-related issues. In nanofluid research, the most commonly used neural network model is Multilayer perceptron neural network (MLP), while the Radial Basis Function Neural Network (RBF-ANN) is the preferred training method. the Generalized Regression Neural Networks (GRNNs) exhibit a simple structure that reduces learning time, making them particularly suitable for nanofluids modelling. Consequently, for nanofluids with a large number of samples, the use of RBF-ANN is recommended. The findings demonstrate the substantial potential of ANN methods as predictive and optimization tools for nanofluids. This paper highlights the recent researches done for evaluating thermo-physical properties of nanofluids using AI algorithms.