Rare earth compound prediction method based on pre-training fine-tuning graph neural network

By using a pre-trained fine-tuned graph neural network, the problem of long prediction time and high cost in traditional rare earth compound prediction methods is solved. This method achieves efficient and accurate screening and prediction of rare earth compounds, provides a clear experimental direction, and helps in the discovery of novel rare earth compounds.

CN121583374APending Publication Date: 2026-02-27UESTC (SHENZHEN) ADVANCED RES INST
0 Cites 0 Cited by

Patent Information

Application Number
CN202511718296.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-11-21
Publication Date
2026-02-27

AI Technical Summary

Technical Problem

Existing technologies are insufficient for efficiently and accurately predicting the stability and feasibility of unknown rare earth compounds. Traditional methods are time-consuming, costly, and have limited coverage. Density functional theory requires prior input of compound structure information, making it difficult to apply.

Method used

A method based on pre-trained fine-tuned graph neural networks is adopted to generate a set of rare earth compounds through permutation and combination, screen combinations that meet chemical rules, construct graph structure data and input it into the pre-trained fine-tuned graph neural network model, output the predicted formation energy and feasibility probability of the compounds, and combine rare earth node weighting and multi-stage screening to achieve efficient screening of compounds.

Benefits of technology

It enables the rapid generation of a set of candidate rare earth compounds that conform to scientific principles and possess innovation, reduces the dependence on experimental structural data, significantly improves the accuracy of formation energy and feasibility prediction, shortens the research and development cycle, and reduces costs.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention belongs to the field of chemical informatization, provides a rare earth compound prediction method based on a pre-training fine-tuning graph neural network, and aims to solve the problem of low prediction efficiency caused by combination explosion and complex properties. According to the main scheme, binary compound combinations of rare earth elements and other elements are generated through permutation and combination, and screening is conducted according to chemical rules and a known compound database to obtain a candidate rare earth compound set; converting each compound in the candidate rare earth compound set into graph structure data comprising node features and edge features, wherein the node features comprise rare earth mark variables used for marking rare earth elements; inputting the graph structure data into a pre-trained fine-tuned graph neural network model to output prediction formation energy and prediction feasibility probability of each candidate rare earth compound; and calculating a comprehensive score of each candidate rare earth compound according to the prediction formation energy and the prediction feasibility probability, and sorting and screening the candidate rare earth compounds according to the comprehensive scores.
Need to check novelty before this filing date? Find Prior Art