Abstract
This paper presents the concept of reduced order machine learning finite element (FE) method. In particular, we propose an example of such method, the proper generalized decomposition (PGD) reduced hierarchical deep-learning neural networks (HiDeNN), called HiDeNN-PGD. We described first the HiDeNN interface seamlessly with the current commercial and open source FE codes. The proposed reduced order method can reduce significantly the degrees of freedom for machine learning and physics based modeling and is able to deal with high dimensional problems. This method is found more accurate than conventional finite element methods with a small portion of degrees of freedom. Different potential applications of the method, including topology optimization, multi-scale and multi-physics material modeling, and additive manufacturing, will be discussed in the paper.
| Original language | English (US) |
|---|---|
| Journal | CMES - Computer Modeling in Engineering and Sciences |
| Volume | 129 |
| Issue number | 1 |
| DOIs | |
| State | Published - 2021 |
Funding
Funding Statement: WKL, YL, HL, SS, SM, AAA are supported by NSF Grants CMMI-1934367 and 1762035. In addition, WKL and SM are supported by AFOSR, USA Grant FA9550-18-1-0381. Acknowledgement: The authors would like to acknowledge the support of the National Science Foundation under Grant Nos. CMMI-1762035 and CMMI-1934367 and AFOSR under Grant No. FA9550-18-1-0381.
Keywords
- Additive manufacturing
- HiDeNN-PGD
- Machine learning
- Model reduction
- Multi-scale modeling
- Topology optimization
ASJC Scopus subject areas
- Software
- Modeling and Simulation
- Computer Science Applications
