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 2409.01949math.NAcs.NA
收藏

ELM-FBPINNs: 一种高效的多级随机特征方法

ELM-FBPINNs: An Efficient Multilevel Random Feature Method

University of Strathclyde(思克莱德大学);Eindhoven University of Technology(埃因霍温理工大学);Imperial College London(伦敦帝国学院)

Samuel Anderson, Victorita Dolean, Ben Moseley, Jennifer Pestana

AI总结:

本文提出了一种结合多级域分解和分区-整体构造的随机特征模型,称为多级ELM-FBPINN,通过将可训练子域网络替换为极端学习机,消除了反向传播,将训练转化为结构化的线性最小二乘问题,从而在多个基准问题中实现了高效的收敛和鲁棒性。

摘要:

领域分解的物理信息神经网络(PINNs)变种,如有限基PINNs(FBPINNs),通过局部化缓解了PINNs的一些问题,如收敛慢和频谱偏差,但仍然依赖于每个子域内的迭代非线性优化。在本文中,我们提出了一种混合方法,将多级域分解和分区-整体构造与随机特征模型相结合,得到一种称为多级ELM-FBPINN的方法。通过将可训练子域网络替换为极端学习机,所得到的公式完全消除了反向传播,并将训练转化为结构化的线性最小二乘问题。我们提供了系统性的数值研究,比较ELM-FBPINNs和多级ELM-FBPINNs与标准PINNs和FBPINNs在代表性基准问题上的性能,证明ELM-FBPINNs和多级ELM-FBPINNs在准确性和收敛速度和对架构和优化参数的鲁棒性方面具有竞争力。通过消融研究,我们进一步阐明了域分解和随机特征丰富在控制表达性、条件和可扩展性中的不同作用。

英文摘要:

Domain-decomposed variants of physics-informed neural networks (PINNs) such as finite basis PINNs (FBPINNs) mitigate some of PINNs' issues like slow convergence and spectral bias through localisation, but still rely on iterative nonlinear optimisation within each subdomain. In this work, we propose a hybrid approach that combines multilevel domain decomposition and partition-of-unity constructions with random feature models, yielding a method referred to as multilevel ELM-FBPINN. By replacing trainable subdomain networks with extreme learning machines, the resulting formulation eliminates backpropagation entirely and reduces training to a structured linear least-squares problem. We provide a systematic numerical study comparing ELM-FBPINNs and multilevel ELM-FBPINNs with standard PINNs and FBPINNs on representative benchmark problems, demonstrating that ELM-FBPINNs and multilevel ELM-FBPINNs achieve competitive accuracy while significantly accelerating convergence and improving robustness with respect to architectural and optimisation parameters. Through ablation studies, we further clarify the distinct roles of domain decomposition and random feature enrichment in controlling expressivity, conditioning, and scalability.

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。

↑