Method and device for predicting properties of metal material based on machine learning force field and medium

The invention relates to the technical field of metal material property prediction, in particular to a method, equipment and medium for predicting metal material properties based on a machine learning force field, which can realize calculation efficiency close to empirical potential, obtain a simulation track through first principle molecular dynamics and an empirical force field and extract a configuration. A database capable of representing the microstructure evolution of the metal material is obtained, so that a metal material machine learning force field for accurately representing the microstructure evolution process of the metal material is established; the problems that a molecular dynamics force field model used for microstructure evolution simulation of the metal material in the hydrogen environment is poor in precision and incomplete in force field are solved, the problem that current metal material molecular dynamics simulation is low in precision is solved, meanwhile, the defect that the first pri