Prediction of Heat Transfer Performance of Heat Pipe Using Machine Learning Approach

The heat pipe is widely used for electronic cooling. ln the simplest configuration of heat pipe, it consists of a cylindrical vessel having three sections (Evaporator, Adiabatic, Condenser). It has a wide range of applications, from smartphones to satellites. Different working fluids are used in the heat pipe based on the operating conditions. This paper discusses the machine learning approach along with some analytical methods to predict the performance of Heat Pipes. Regression, Random Forest, Support Vector Machine (SVM), and Gaussian Matrix Model (GMM) have been tested to find an optimal model based on 2500 experimental data points collected from different research papers and to show the relationship between different parameters. Our objective is to predict the heat transfer performance of a heat pipe by using different machine learning algorithms.