|
|
ESR-PINNs: Physics-informed neural networks with expansion-shrinkage resampling selection strategies |
Jianan Liu(刘佳楠)1, Qingzhi Hou(侯庆志)2,†, Jianguo Wei(魏建国)1, and Zewei Sun(孙泽玮)1 |
1 College of Intelligence and Computing, Tianjin University, Tianjin 300350, China; 2 State Key Laboratory of Hydraulic Engineering Simulation and Safety, Tianjin University, Tianjin 300350, China |
|
|
Abstract Neural network methods have been widely used in many fields of scientific research with the rapid increase of computing power. The physics-informed neural networks (PINNs) have received much attention as a major breakthrough in solving partial differential equations using neural networks. In this paper, a resampling technique based on the expansion-shrinkage point (ESP) selection strategy is developed to dynamically modify the distribution of training points in accordance with the performance of the neural networks. In this new approach both training sites with slight changes in residual values and training points with large residuals are taken into account. In order to make the distribution of training points more uniform, the concept of continuity is further introduced and incorporated. This method successfully addresses the issue that the neural network becomes ill or even crashes due to the extensive alteration of training point distribution. The effectiveness of the improved physics-informed neural networks with expansion-shrinkage resampling is demonstrated through a series of numerical experiments.
|
Received: 28 December 2022
Revised: 28 February 2023
Accepted manuscript online: 07 March 2023
|
PACS:
|
07.05.Mh
|
(Neural networks, fuzzy logic, artificial intelligence)
|
|
02.30.Jr
|
(Partial differential equations)
|
|
84.35.+i
|
(Neural networks)
|
|
Fund: Project supported by the National Key Research and Development Program of China (Grant No. 2020YFC1807905), the National Natural Science Foundation of China (Grant Nos. 52079090 and U20A20316), and the Basic Research Program of Qinghai Province (Grant No. 2022-ZJ-704). |
Corresponding Authors:
Qingzhi Hou
E-mail: qhou@tju.edu.cn
|
Cite this article:
Jianan Liu(刘佳楠), Qingzhi Hou(侯庆志), Jianguo Wei(魏建国), and Zewei Sun(孙泽玮) ESR-PINNs: Physics-informed neural networks with expansion-shrinkage resampling selection strategies 2023 Chin. Phys. B 32 070702
|
[1] Chen W, Guo Y and Jin S W 2020 Acta Phys. Sin. 69 240502 (in Chinese) [2] Xu Q W, Zheng Z and Jiang H B 2022 Chin. Phys. B 31 024302 [3] Radford A, Wu J, Child R, Luan D, Amodei D and Sutskever I 2014 OpenAI blog 1 9 [4] Yang Z, Dai Z, Yang Y, Carbonell J, R.Salakhutdinov R and Le Q V 2019 32nd Conference on Neural Information Processing Systems, December 8-14, 2019, Vancouver, Canada, pp. 5753-5763 [5] Snyder D, Garcia-Romero D, Sell G, Povey D and Khudanpur S 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), April 15-20, 2018, Calgary, AB, Canada, pp. 5329-5333 [6] He Y, Sainath T N, Prabhavalkar R et al. 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), May 12-17, 2019, Brighton, UK, pp. 6381-6385 [7] Han J, Jentzen A and Weinan E 2018 Proc. Natl. Acad. Sci. USA 115 8505 [8] Weinan E 2021 Not. Am. Math. Soc. 68 565 [9] Carleo G, Cirac I, Cranmer K, Daudet L, Schuld M, Tishby N, Vogt-Maranto L and Zdeborová L 2019 Rev. Mod. Phys. 91 045002 [10] Zhou Z Z and Yan Z Y 2021 Commun. Theor. Phys. 73 105006 [11] Raissi M, Perdikaris P and Karniadakis G E 2017 arXiv:1711.10561 [cs.AI] [12] Cai S, Wang Z, Wang S, Perdikaris P and Karniadakis G E 2021 J. Heat Transfer 143 060801 [13] Yin M, Zheng X, Humphrey J D and Karniadakis G E 2021 Comput. Methods Appl. Mech. Eng. 375 113603 [14] Chen Y, Lu L, Karniadakis G E and Dal N L 2020 Opt. Express 28 11618 [15] Cai S, Wang Z, Fuest F, Jeon Y J, Gray C and Karniadakis G E 2021 J. Fluid Mech. 915 A102 [16] Jin X, Cai S, Li H and Karniadakis G E 2021 J. Comput. Phys. 426 109951 [17] Raissi M, Yazdani A and Karniadakis G E 2020 Science 367 1026 [18] Fang B L, Wang J G and Feng G B 2022 Acta Phys. Sin. 71 200601 (in Chinese) [19] Daw A, Bu J, Wang S, Perdikaris P and Karpatne A 2022 arXiv:2207.02338 [cs.LG] [20] Nabian M A, Gladstone R J and Meidani H 2021 Computer-Aided Civil and Infrastructure Engineering 36 962 [21] Bischof R and Kraus M 2021 arXiv:2110.09813 [cs.LG] [22] Wang S, Teng Y and Perdikaris P 2021 SIAM J. Sci. Comput. 43 A3055 [23] Lyu L, Zhang Z, Chen M and Chen J 2022 J. Comput. Phys. 452 110930 [24] Li W, Zhang C, Wang C, Guan H and Tao D 2022 arXiv:2205.08754 [cs.LG] [25] McClenny L and Braga-Neto U 2020 arXiv:2009.04544 [cs.LG] [26] Wight C L and Zhao J 2020 arXiv:2007.04542 [cs.LG] [27] Mattey R and Ghosh S 2022 Comput. Methods Appl. Mech. Eng. 390 114474 [28] Tian S F and Li B 2023 Acta Phys. Sin. 72 100202 (in Chinese) [29] Lu L, Meng X, Mao Z and Karniadakis G E 2021 SIAM Rev. 63 208 [30] Beneš M, Chalupecký V and Mikula K 2004 Appl. Numer. Math. 51 187 [31] Ilmanen T 1993 J. Diff. Geometry 38 417 [32] Lu L, Pestourie R, Yao W, Wang Z, Verdugo F and Johnson S G 2021 SIAM J. Sci. Comput. 43 B1105 [33] Xu Z J, Zhang Y and Xiao Y 2019 International Conference on Neural Information Processing (ICONIP 2019). Lecture Notes in Computer Science Vol. 11953 (Cham: Springer) [34] Ghia U, Ghia K N and Shin C 1982 J. Comput. Phys. 48 387 |
No Suggested Reading articles found! |
|
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
Altmetric
|
blogs
Facebook pages
Wikipedia page
Google+ users
|
Online attention
Altmetric calculates a score based on the online attention an article receives. Each coloured thread in the circle represents a different type of online attention. The number in the centre is the Altmetric score. Social media and mainstream news media are the main sources that calculate the score. Reference managers such as Mendeley are also tracked but do not contribute to the score. Older articles often score higher because they have had more time to get noticed. To account for this, Altmetric has included the context data for other articles of a similar age.
View more on Altmetrics
|
|
|