Self-adaptive physics-informed neural networks

人工神经网络 计算机科学 统计物理学 人工智能 理论物理学 物理
作者
Levi D. McClenny,Ulisses Braga-Neto
出处
期刊:Journal of Computational Physics [Elsevier BV]
卷期号:474: 111722-111722 被引量:426
标识
DOI:10.1016/j.jcp.2022.111722
摘要

Physics-Informed Neural Networks (PINNs) have emerged recently as a promising application of deep neural networks to the numerical solution of nonlinear partial differential equations (PDEs). However, it has been recognized that adaptive procedures are needed to force the neural network to fit accurately the stubborn spots in the solution of “stiff” PDEs. In this paper, we propose a fundamentally new way to train PINNs adaptively, where the adaptation weights are fully trainable and applied to each training point individually, so the neural network learns autonomously which regions of the solution are difficult and is forced to focus on them. The self-adaptation weights specify a soft multiplicative soft attention mask, which is reminiscent of similar mechanisms used in computer vision. The basic idea behind these SA-PINNs is to make the weights increase as the corresponding losses increase, which is accomplished by training the network to simultaneously minimize the losses and maximize the weights. In addition, we show how to build a continuous map of self-adaptive weights using Gaussian Process regression, which allows the use of stochastic gradient descent in problems where conventional gradient descent is not enough to produce accurate solutions. Finally, we derive the Neural Tangent Kernel matrix for SA-PINNs and use it to obtain a heuristic understanding of the effect of the self-adaptive weights on the dynamics of training in the limiting case of infinitely-wide PINNs, which suggests that SA-PINNs work by producing a smooth equalization of the eigenvalues of the NTK matrix corresponding to the different loss terms. In numerical experiments with several linear and nonlinear benchmark problems, the SA-PINN outperformed other state-of-the-art PINN algorithm in L2 error, while using a smaller number of training epochs.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
sky完成签到,获得积分10
刚刚
Sue完成签到,获得积分10
1秒前
1秒前
唐水之完成签到,获得积分10
1秒前
NQP发布了新的文献求助10
1秒前
threewei完成签到,获得积分10
1秒前
1秒前
eason楽发布了新的文献求助10
1秒前
2秒前
无极微光的应助被awang采纳,获得20
2秒前
2秒前
半夏发布了新的文献求助10
3秒前
tang发布了新的文献求助10
3秒前
轻松不正发布了新的文献求助10
4秒前
4秒前
inn完成签到,获得积分10
4秒前
4秒前
OoO发布了新的文献求助10
4秒前
5秒前
大刚完成签到,获得积分10
5秒前
宁静致远发布了新的文献求助10
5秒前
英姑的应助被obito采纳,获得10
5秒前
crescendo完成签到,获得积分10
6秒前
美丽的周完成签到,获得积分10
6秒前
6秒前
温暖砖头发布了新的文献求助10
7秒前
李想完成签到,获得积分10
7秒前
科研通AI65的应助被鸭梨山大33采纳,获得10
7秒前
科研通AI6.4的应助被周一采纳,获得10
8秒前
顺意发布了新的文献求助10
8秒前
独白发布了新的文献求助10
9秒前
www发布了新的文献求助10
9秒前
9秒前
9秒前
饱满小夏完成签到,获得积分10
9秒前
10秒前
10秒前
10秒前
wanluxia完成签到,获得积分10
10秒前
景平完成签到,获得积分0
10秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
八旗制度の研究 600
Improving the wear resistance of vehicle gear oil by oil soluble LaF3 nanoparticles and typical additives of gear oil 500
Boundaries and Bodhisattvas 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8183046
求助须知:如何正确求助?哪些是违规求助? 9572370
关于积分的说明 21824411
捐赠科研通 7779694
什么是DOI,文献DOI怎么找? 3433792
关于科研通互助平台的介绍 2558024
邀请新用户注册赠送积分活动 2479971