Related Experiment Video
Updated: May 4, 2026
06:00
Electroantennography-based Bio-hybrid Odor-detecting Drone using Silkmoth Antennae for Odor Source Localization
Published on: August 27, 2021
5.1K
Deep-learning approach for developing bilayered electromagnetic interference shielding composite aerogels based on
Chenglei He1, Liya Yu1, Yun Jiang1
1College of Mechanical Engineering, Guizhou University, Guiyang 550025, China.
Journal of Colloid and Interface Science
|February 23, 2025
Summary
A new multimodal neural network model accurately predicts electromagnetic interference (EMI) shielding performance in advanced aerogel materials. This approach accelerates the development of high-performance EMI shielding materials, reducing experimental costs and time.
Area of Science:
- Materials Science
- Nanotechnology
- Artificial Intelligence
Background:
- Developing high-performance electromagnetic interference (EMI) shielding materials requires cost-effective and efficient methods.
- Traditional experimental approaches are time-consuming and resource-intensive.
- Predictive modeling offers a promising alternative for accelerating material discovery.
Purpose of the Study:
- To propose a multimodal data fusion neural network model for predicting EMI shielding performance.
- To evaluate the model's accuracy and generalization capability for silver-modified four-pronged zinc oxide/waterborne polyurethane/barium ferrite (Ag@F-ZnO/WPU/BF) aerogels.
- To demonstrate a non-experimental approach for developing advanced EMI shielding materials.
Main Methods:
- Preparation of 16 Ag@F-ZnO/WPU/BF aerogel samples with varying compositions using pre-casting and directional freezing.
- Development of a multimodal neural network model integrating composite ingredients and microstructural images.
- Utilizing a combination of fully connected neural network (FCNN) and residual neural network (ResNet) with GatedFusion for prediction.
Main Results:
- Prepared aerogels exhibited excellent EMI shielding effectiveness (SET) up to 78.6 dB and absorption coefficient of 0.96.
- The multimodal FCNN-ResNet model achieved high prediction accuracy with RMSE of 0.7626, MAE of 0.4918, and R of 0.9885.
- The model demonstrated strong generalization, predicting EMI performance of new aerogels with less than 5% average error.
Conclusions:
- The multimodal neural network model effectively predicts EMI shielding performance of Ag@F-ZnO/WPU/BF aerogels.
- This data-driven approach significantly improves accuracy and efficiency in material property prediction.
- The study offers a pathway to reduce experimental burdens and accelerate the development of novel EMI shielding materials.
