期刊
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
卷 59, 期 3, 页码 787-799出版社
SPRINGER
DOI: 10.1007/s00158-018-2101-5
关键词
Deep learning; Machine learning; Topology optimization; Generative model; Generative adversarial network; Convolutional neural network
资金
- National Research Foundation of Korea (NRF) - Korea government (MSIT) [NRF-2018R1C1B6005157]
- National Institute of Supercomputing and Network (NISN)/Korea Institute of Science and Technology Information (KISTI) [KSC-2017-S1-0029]
In this study, we propose a novel deep learning-based method to predict an optimized structure for a given boundary condition and optimization setting without using any iterative scheme. For this purpose, first, using open-source topology optimization code, datasets of the optimized structures paired with the corresponding information on boundary conditions and optimization settings are generated at low (32 x 32) and high (128 x 128) resolutions. To construct the artificial neural network for the proposed method, a convolutional neural network (CNN)-based encoder and decoder network is trained using the training dataset generated at low resolution. Then, as a two-stage refinement, the conditional generative adversarial network (cGAN) is trained with the optimized structures paired at both low and high resolutions and is connected to the trained CNN-based encoder and decoder network. The performance evaluation results of the integrated network demonstrate that the proposed method can determine a near-optimal structure in terms of pixel values and compliance with negligible computational time.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据