A transformer gas-sensitive material screening method and device and a storage medium

By constructing a dual-model architecture that combines a global prediction model and an engineering sub-model, the path dependence and resource waste problems in material screening for transformer fault gas detection are solved, enabling efficient and reliable screening and verification of gas-sensitive materials, and improving the accuracy and R&D efficiency of transformer fault detection.

CN122290782BActive Publication Date: 2026-07-21CHONGQING UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
CHONGQING UNIV
Filing Date
2026-05-26
Publication Date
2026-07-21

AI Technical Summary

Technical Problem

In the detection of fault gases in transformers, existing technologies for screening gas-sensitive materials struggle to balance the overall understanding of the material space with the needs of engineering applications, resulting in path dependence and resource waste.

Method used

A dual-model architecture combining a global prediction model and an engineering sub-model is constructed. Through feature contribution analysis and engineering rule constraints, hierarchical scheduling of material exploration and verification is achieved, reducing path dependency risk and improving the interpretability and engineering applicability of screening results.

Benefits of technology

This effectively reduces path dependence risks in the material screening process, enhances the integrity of material spatial cognition and the engineering reliability of screening results, controls the input of verification resources, and improves the R&D efficiency and detection accuracy of transformer fault gas detection materials.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122290782B_ABST
    Figure CN122290782B_ABST
Patent Text Reader

Abstract

The application discloses a transformer gas-sensitive material screening method and device and a storage medium, and belongs to the technical field of power equipment state monitoring and gas sensing material design. The method comprises the following steps: obtaining original characteristic data of a candidate gas-sensitive material to be screened; extracting corresponding characteristic values from the original characteristic data according to a pre-determined characteristic subset, wherein the characteristic subset is obtained through characteristic contribution analysis based on a pre-constructed global prediction model; inputting the extracted characteristic values into a pre-trained engineering sub-model, combining engineering rule constraints to obtain a comprehensive score of the candidate material; dividing the candidate material into engineering levels according to the comprehensive score, and outputting a material list meeting engineering access conditions. The application realizes engineering control of the material exploration process, and improves the reliability and engineering applicability of the material screening process.
Need to check novelty before this filing date? Find Prior Art