arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2401.02810cs.LGcs.AIcs.NAmath.NA
收藏

使用迁移学习的物理信息神经网络求解高频与多尺度问题

Physics-Informed Neural Networks for High-Frequency and Multi-Scale Problems using Transfer Learning

University of Genoa(热那亚大学);Kiel University(基尔大学);University of Waterloo(滑铁卢大学);GeoAnalysis Engineering GmbH(GeoAnalysis Engineering 有限公司)

Abdul Hannan Mustajab, Hao Lyu, Zarghaam Rizvi, Frank Wuttke

AI总结:

本文提出用迁移学习从低频到高频逐步训练物理信息神经网络,以提升求解高频多尺度问题时的鲁棒性与收敛性,并验证其可在不增加参数的情况下减少数据需求与训练时间。

摘要:

物理信息神经网络(PINN)是一种用于偏微分方程和常微分方程(ODEs/PDEs)的数据驱动求解器。它提供了一个统一框架来处理正问题和逆问题。然而,目标函数的复杂性往往导致训练失败。这一问题在求解高频和多尺度问题时尤为突出。我们提出使用迁移学习来提高PINN训练的鲁棒性和收敛性,从低频问题开始训练,并逐步逼近高频问题。通过两个案例研究,我们发现迁移学习能够有效地训练PINN,在不增加网络参数的情况下,从低频问题到高频问题逼近解。此外,它需要更少的数据点和更短的训练时间。我们详细描述了训练策略,包括优化器选择,并给出了使用迁移学习训练神经网络以求解更复杂问题的指导原则。

英文摘要:

Physics-informed neural network (PINN) is a data-driven solver for partial and ordinary differential equations(ODEs/PDEs). It provides a unified framework to address both forward and inverse problems. However, the complexity of the objective function often leads to training failures. This issue is particularly prominent when solving high-frequency and multi-scale problems. We proposed using transfer learning to boost the robustness and convergence of training PINN, starting training from low-frequency problems and gradually approaching high-frequency problems. Through two case studies, we discovered that transfer learning can effectively train PINN to approximate solutions from low-frequency problems to high-frequency problems without increasing network parameters. Furthermore, it requires fewer data points and less training time. We elaborately described our training strategy, including optimizer selection, and suggested guidelines for using transfer learning to train neural networks for solving more complex problems.

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。

↑