A solid waste-based building material proportioning design method and device based on multi-objective deep learning

By employing a multi-objective deep learning approach, combined with multi-modal data and multi-objective optimization algorithms, the problems of low efficiency, insufficient accuracy, and multi-objective optimization in the design of multi-source solid waste raw material proportions were solved. This resulted in efficient and accurate building material proportion design, improving the R&D efficiency and performance of solid waste-based building materials.

CN122436075APending Publication Date: 2026-07-21HENAN BUILDING MATERIALS RES & DESIGN LNSTITUTE CO LTD +2
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
HENAN BUILDING MATERIALS RES & DESIGN LNSTITUTE CO LTD
Filing Date
2026-04-09
Publication Date
2026-07-21

AI Technical Summary

Technical Problem

Existing technologies for processing multi-source solid waste materials suffer from low efficiency in proportioning design, high cost, insufficient prediction accuracy, and difficulty in multi-objective collaborative optimization, failing to meet the building materials industry's demand for high efficiency, high performance, and green utilization.

Method used

This paper employs a multi-objective deep learning-based approach. By establishing a multi-modal characteristic database of multi-source solid waste, a hierarchical deep learning prediction model combined with a multi-objective optimization algorithm is used to achieve high-precision multi-objective performance prediction and global proportion optimization. The method integrates macroscopic physicochemical data, microscopic morphological image data, and mineral phase composition spectral data. It utilizes a hierarchical graph attention fusion network (HGAF-Net) and a constrained adaptive genetic algorithm (CA-NSGA-II) for in-depth mining of complex nonlinear relationships and multi-objective optimization.

Benefits of technology

It has shortened the traditional method of proportion design, which takes months, to a few hours through automated calculation, improved prediction accuracy and multi-objective optimization capabilities, and provided a series of proportion schemes that achieve optimal balance in multiple dimensions such as performance, cost, and environmental protection, thereby improving the R&D cycle and resource utilization rate of solid waste-based building materials.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122436075A_ABST
    Figure CN122436075A_ABST
Patent Text Reader

Abstract

The application discloses a kind of solid waste infrastructure material proportioning design method and device based on multi-objective deep learning, comprising: first, build a multi-source solid waste multi-modal characteristic database including macroscopic physicochemical data, microscopic topography image data and mineral phase composition spectrum data;Second, the multi-modal data representing a proportioning scheme is input into a pre-trained hierarchical graph attention fusion network (HGAF-Net) model to simultaneously predict multiple key performance indicators of building materials;Finally, the prediction model is used as a fitness function, and a global optimization is performed on multiple design objectives through a constrained adaptive genetic algorithm (CA-NSGA-II) to generate a set of Pareto optimal proportioning scheme set.The application also provides a device for implementing the method.The application solves the problems of low efficiency, high cost, difficulty in handling complex interactions of multi-source solid waste and multi-objective collaborative optimization of traditional proportioning design methods, and provides an efficient, accurate and green intelligent decision support tool.
Need to check novelty before this filing date? Find Prior Art