The 11th International Conference on Fuzzy Systems and Data Mining (FSDM 2025)

Abstract: This report proposes the development and implementation of an innovative machine learning surrogate meshfree approach for modelling heat transfer phenomena in industrial processes with moving heat sources. This approach is based on the Finite Pointset Method (FPM), a Generalized Finite Differences Method (GFDM) which is enhanced in this work with machine learning techniques for the construction of shape functions. The objective of this development is to overcome the limitations posed by commonly used mesh-based methods and to improve the efficiency found in standard mesh-free methods. The tool resulting from this work contributes to the positive impact of mathematical modelling developments with mesh-free methods and expands their applications in areas of science and industry. The accuracy, feasibility, and robustness of this new formulation are evaluated in detail through a series of case studies, starting with simple cases, for which analytical solutions exist, and evolving to complex and practical examples, representative of selected applications involving moving heat sources. The results of these examples demonstrate that this meshless numerical tool is promising, and furthermore, similar strategies could be proposed for modelling other physical phenomena or their coupling.

Biography: Dr. Felix Raymundo Saucedo Zendejo obtained his doctorate in Engineering Sciences at the National Technological Institute of Mexico conducting research with a specialty in Computational Fluid Mechanics, Numerical Methods, Mesh-Free Methods and Simulation of industrial processes. He has a master's degree in materials science from the National Technological Institute of Mexico and has a bachelor's degree in physics from the Faculty of Physico-Mathematical Sciences of the Autonomous University of Coahuila. Since October 2024 he is the head of the Research Center for Applied Mathematics from the Autonomous University of Coahuila. Over the past years, Dr. Felix Raymundo Saucedo Zendejo has been working on meshfree methods developing for the analysis and modeling of different engineering processes and physical phenomena. Results obtained on his research have been published in several international journals indexed in JCR/SCOPUS with impact factor and some of them have been cited in several research articles. He has collaborated on research projects with international partners from the University of Minho in Portugal, Perm National Research Polytechnic University in Russia, Silesian University of Technology in Poland, and Suranaree University of Technology in Thailand. He has extensive experience as a teacher in undergraduate lectures such as Classical Mechanics, Transport Phenomena, Electricity and Magnetism, Simulation, and Computational Physics, as well as in postgraduate lectures such as Applied Mathematics, Programming, and Computational Physics, among others.