Publication

Abstract : The prediction of heat transmission will be beneficial in the areas of frost protection, subterranean cables, slab thermal performance, and soil temperature management, all of which include the use of a semi-infinite medium. The conduction of heat through a semi-infinite medium has now become a prominent problem and can be analyzed mathematically through traditional methods. For describing the distribution of temperature, many analytical methods, together with machine learning (ML) techniques that will enhance the efficiency and accuracy of the scenario, have been implemented. Following this, an ML technique based on the Fourier transform method is proposed here for the investigation of heat conduction in a semi-infinite medium. In accordance with the literature survey conducted, no prior study is juxtaposed based on these two methodologies. The Fourier transform method serves precisely for solving heat conduction problems, while ML techniques provide a cutting-edge, data-driven method that could identify nonlinearities and intricate patterns in thermal management. Numerous ML techniques are employed for analyzing the heat conduction problem and are compared with the solution of the Fourier transform method. From the evaluation metrics employed, the unit R-squared (R2) value, mean squared error (MSE), root mean square error (RMSE), and mean absolute error (MAE) values approaching 0 demonstrate that the decision tree model outperforms other ML techniques better and is highly predictive for heat conduction problems.