arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2008.12649cs.LGphysics.app-ph
收藏

偏微分方程深度代理的主动学习:在元表面设计中的应用

Active learning of deep surrogates for PDEs: Application to metasurface design

MIT(麻省理工学院);IBM Research AI(IBM研究院人工智能部门);IBM Thomas J Watson Research Center(IBM托马斯·J·沃森研究院);MIT-IBM Watson AI Lab(麻省理工学院-IBM沃森人工智能实验室);Iowa State University(爱荷华州立大学)

Raphaël Pestourie, Youssef Mroueh, Thanh V. Nguyen, Payel Das, Steven G. Johnson

摘要:

偏微分方程的代理模型广泛应用于超材料设计中,以快速评估可组合部件的行为。然而,通过机器学习训练精确代理的训练成本会随着变量的数量而迅速增加。对于光子器件模型,我们发现,这种训练变得特别具有挑战性的设计区域增长大于光波长。我们提出了一个主动学习算法,减少了一个数量级以上的神经网络代理模型的光学表面组件相比,随机样本的训练点的数量。结果表明,代理评估是超过两个数量级的速度比直接解决,我们演示了如何利用这一点来加速大规模的工程优化。

英文摘要:

Surrogate models for partial-differential equations are widely used in the design of meta-materials to rapidly evaluate the behavior of composable components. However, the training cost of accurate surrogates by machine learning can rapidly increase with the number of variables. For photonic-device models, we find that this training becomes especially challenging as design regions grow larger than the optical wavelength. We present an active learning algorithm that reduces the number of training points by more than an order of magnitude for a neural-network surrogate model of optical-surface components compared to random samples. Results show that the surrogate evaluation is over two orders of magnitude faster than a direct solve, and we demonstrate how this can be exploited to accelerate large-scale engineering optimization.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑