Progress in the Application of Mathematical Algorithms in Materials Science

Abstract

As materials are the direct drive for scientific development and social progress, the research on their structure and properties has always been a research hotspot in materials science. At present, the discovery of new materials is mainly carried out by the researcher's intuitive judgment of materials and a large number of "trial and error" experiments, which is inefficient and difficult to effectively discover a large number of possible new material combinations. Materials science has long relied on two major means of experiment and calculation. The machine learning-based mathematical algorithms offers a new research method. This article systematically expounds the common mathematical algorithms and the basic processes of their calculation methods, details the application of mathematical algorithms in material performance prediction and new material design, and gives a summarization and outlook of the challenges and application prospects concerning their application in materials science.