Model construction method and aging performance prediction method, electronic device, and storage medium

By constructing a micro-macro correlation model and combining machine learning and interpretability analysis, the problems of accuracy and interpretability in predicting the aging performance of bamboo reconstituted materials were solved, providing high-precision aging performance prediction and modification design guidance.

CN121812036BActive Publication Date: 2026-05-29INST OF WOOD INDUDTRY CHINESE ACAD OF FORESTRY
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202610290155.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-03-11
Publication Date
2026-05-29
Estimated Expiration
2046-03-11

Smart Images

  • Figure CN121812036B_ABST
    Figure CN121812036B_ABST
Patent Text Reader

Abstract

The present disclosure provides a model construction method and an aging performance prediction method, an electronic device and a storage medium, and relates to the field of biomass composite material performance prediction. The model construction method comprises: obtaining multi-scale aging characteristic data samples under a target service climate condition, including environmental meteorological data samples, and micro-chemical characteristic data samples and macro-mechanical performance data samples of biomass composite materials under the action of the environmental meteorological data samples; constructing a micro-macro correlation model based on the micro-chemical characteristic data samples and the macro-mechanical performance data samples; screening key micro-feature samples from the micro-chemical characteristic data samples based on an explainability analysis method; for each key micro-feature sample, a time series prediction model is constructed respectively with the environmental meteorological data samples as exogenous variables; and cascading the time series prediction model corresponding to each key micro-feature sample and the micro-macro correlation model to obtain an aging performance prediction model of the biomass composite material.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Biomass composite material aging performance prediction method based on neural network

    CN117497104A