一种淀粉水凝胶微观结构逆向设计方法
By constructing a physical property feature mapping encoder, a multi-scale self-attention microstructure generator, and a dual-stream physical consistency discriminator, the nonlinear mapping problem between microstructure and macroscopic properties in starch hydrogel research and development was solved, enabling rapid and accurate microstructure design and performance prediction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- OCEAN UNIV OF CHINA
- Filing Date
- 2026-01-05
- Publication Date
- 2026-07-17
AI Technical Summary
The existing research and development model for starch hydrogels relies on the traditional forward trial-and-error method, which results in long research and development cycles, high resource consumption, and difficulty in accurately quantifying the nonlinear relationship between microstructure and macroscopic performance. It is also impossible to directly deduce the ideal microstructure from the target performance indicators.
A reverse design method for starch hydrogel microstructures is adopted. By constructing a physical property feature mapping encoder, a multi-scale self-attention microstructure generator, and a two-stream physical consistency discriminator, combined with a differentiable physical performance prediction surrogate model, the reverse generation and evaluation of microstructures from target performance indicators are realized.
It enables rapid and precise microstructure design, reduces R&D costs, ensures that the generated starch hydrogel microstructure closely approximates the preset mechanical performance target, and provides scientific guidance for process parameters.
Smart Images
Figure CN121862281B_ABST
Abstract
Citation Information
Patent Citations
Reverse design method of novel super-fiber reinforced hydrogel structure based on GAN
CN118053526A
Pneumatic deformation interface PSCI profiling control method and device based on machine learning
CN118734661A