|
| Current and future trends of artificial intelligence in the field of structural topology optimization |
| Received:May 15, 2021 Revised:June 08, 2021 |
| View Full Text View/Add Comment Download reader |
| DOI：10.7511/jslx20210517401 |
| KeyWord:structural topology optimization artificial intelligence machine learning deep learning |
| Author | Institution |
| 阎军 |
大连理工大学 工业装备结构分析国家重点实验室, 工程力学系, 大连 |
| 许琦 |
大连理工大学 工业装备结构分析国家重点实验室, 工程力学系, 大连 |
| 张起 |
大连理工大学 工业装备结构分析国家重点实验室, 工程力学系, 大连 |
| 范志瑞 |
大连理工大学 工业装备结构分析国家重点实验室, 工程力学系, 大连 |
| 杜洪泽 |
大连理工大学 工业装备结构分析国家重点实验室, 工程力学系, 大连 |
| 耿东岭 |
大连理工大学 工业装备结构分析国家重点实验室, 工程力学系, 大连 |
| 阎琨 |
大连理工大学 化工学院, 大连 |
| 牛斌 |
大连理工大学 机械工程学院, 大连 |
|
| Hits: 1645 |
| Download times: 587 |
| Abstract: |
| Structural optimization, especially structural topology optimization, has received extensive attention from academia and industry. Through the development of different topology optimization algorithms, many innovative designs of optimal topology configurations with excellent multi-disciplinary performance such as in mechanical science, thermal science, and acoustics have been realized. However, traditional topology optimization methods often require thousands of iterative steps when dealing with large-scale topology optimization problems, and face the challenge of high computational complexity due to large-scale finite element analyses. In recent years with the rapid development of artificial intelligence methods represented by machine learning, artificial intelligence-based topology optimization has become the most promising new research direction. By combining artificial intelligence algorithms with the topology optimization framework, the efficiency of structural topology optimization is greatly improved, and at the same time, it is possible to achieve a real-time topology optimization design. This paper reviews some major advances in the research of topology optimization methods based on machine learning in the past ten years, and briefly introduces the state of the art up to now. Due to the limited space of the paper, this review does not involve the complete literature in this field, and its review scope is limited closely to the author's own research interests. |
|
|
|