Graph dynamical networks for forecasting collective behavior of active matter

doi:10.1088/1674-1056/ac7cce

中国物理B ›› 2022, Vol. 31 ›› Issue (11): 116401-116401.doi: 10.1088/1674-1056/ac7cce

Graph dynamical networks for forecasting collective behavior of active matter

Yanjun Liu(刘彦君)¹, Rui Wang(王瑞)¹, Cai Zhao(赵偲)^2,†, and Wen Zheng(郑文)^1,3,‡

1 Institute of Public-Safety and Big Data, College of Data Science, Taiyuan University of Technology, Taiyuan 030060, China;
2 Center of Information Management and Development, Taiyuan University of Technology, Taiyuan 030060, China;
3 Center for Healthy Big Data, Changzhi Medical College, Changzhi 046000, China

收稿日期:2022-02-18 修回日期:2022-06-13 接受日期:2022-06-29 出版日期:2022-10-17 发布日期:2022-10-17
通讯作者: Cai Zhao, Wen Zheng E-mail:zhaocai@tyut.edu.cn;zhengwen@tyut.edu.cn

Graph dynamical networks for forecasting collective behavior of active matter

Yanjun Liu(刘彦君)¹, Rui Wang(王瑞)¹, Cai Zhao(赵偲)^2,†, and Wen Zheng(郑文)^1,3,‡

1 Institute of Public-Safety and Big Data, College of Data Science, Taiyuan University of Technology, Taiyuan 030060, China;
2 Center of Information Management and Development, Taiyuan University of Technology, Taiyuan 030060, China;
3 Center for Healthy Big Data, Changzhi Medical College, Changzhi 046000, China

Received:2022-02-18 Revised:2022-06-13 Accepted:2022-06-29 Online:2022-10-17 Published:2022-10-17
Contact: Cai Zhao, Wen Zheng E-mail:zhaocai@tyut.edu.cn;zhengwen@tyut.edu.cn

摘要/Abstract

摘要： After decades of theoretical studies, the rich phase states of active matter and cluster kinetic processes are still of research interest. How to efficiently calculate the dynamical processes under their complex conditions becomes an open problem. Recently, machine learning methods have been proposed to predict the degree of coherence of active matter systems. In this way, the phase transition process of the system is quantified and studied. In this paper, we use graph network as a powerful model to determine the evolution of active matter with variable individual velocities solely based on the initial position and state of the particles. The graph network accurately predicts the order parameters of the system in different scale models with different individual velocities, noise and density to effectively evaluate the effect of diverse condition. Compared with the classical physical deduction method, we demonstrate that graph network prediction is excellent, which could save significantly computing resources and time. In addition to active matter, our method can be applied widely to other large-scale physical systems.

关键词: active matter, graph network, improvement of Vicsek, collective motion

Abstract: After decades of theoretical studies, the rich phase states of active matter and cluster kinetic processes are still of research interest. How to efficiently calculate the dynamical processes under their complex conditions becomes an open problem. Recently, machine learning methods have been proposed to predict the degree of coherence of active matter systems. In this way, the phase transition process of the system is quantified and studied. In this paper, we use graph network as a powerful model to determine the evolution of active matter with variable individual velocities solely based on the initial position and state of the particles. The graph network accurately predicts the order parameters of the system in different scale models with different individual velocities, noise and density to effectively evaluate the effect of diverse condition. Compared with the classical physical deduction method, we demonstrate that graph network prediction is excellent, which could save significantly computing resources and time. In addition to active matter, our method can be applied widely to other large-scale physical systems.

Key words: active matter, graph network, improvement of Vicsek, collective motion

中图分类号: (Phase equilibria)

64.75.-g

07.05.Mh (Neural networks, fuzzy logic, artificial intelligence) 05.65.+b (Self-organized systems) 87.64.Aa (Computer simulation)

Yanjun Liu(刘彦君), Rui Wang(王瑞), Cai Zhao(赵偲), and Wen Zheng(郑文). Graph dynamical networks for forecasting collective behavior of active matter[J]. 中国物理B, 2022, 31(11): 116401-116401.

Yanjun Liu(刘彦君), Rui Wang(王瑞), Cai Zhao(赵偲), and Wen Zheng(郑文). Graph dynamical networks for forecasting collective behavior of active matter[J]. Chin. Phys. B, 2022, 31(11): 116401-116401.

参考文献

[1] Genkin M M, Sokolov A, Lavrentovich O D and Aranson I S 2017 Phys. Rev. X 7 011029
[2] Jolles J W, Boogert N J, Sridhar V H, Couzin I D and Manica A 2017 Current Biology 27 2862
[3] Bajec I L and Heppner F H 2009 Animal Behaviour 78 777
[4] Dombrowski C, Cisneros L, Chatkaew S, Goldstein R E and Kessler J O 2004 Phys. Rev. Lett. 93 098103
[5] Kumar S, Singh J P, Giri D and Mishra S 2021 Phys. Rev. E 104 024601
[6] Cichos F, Gustavsson K, Mehlig B and Volpe G 2020 Nat. Mach. Intell. 2 94
[7] Keta Y E, Fodor van Wijland F, Cates M E and Jack R L 2021 Phys. Rev. E 103 022603
[8] Zhang J, Zheng W, Zhang S, Xu D, Nie Y, Jiang Z and Xu N 2021 Sci. Adv. 7 eabg6766
[9] Speck T 2020 Soft Matter 16 2652
[10] Chepizhko O, Saintillan D and Peruani F 2021 Soft Matter 17 3113
[11] Zhang J, Zhao Y, Tian B, Peng L, Zhang H T, Wang B H and Zhou T 2009 Physica A 388 1237
[12] Chen D 2018 Research on Evolutionary Analysis and Pattern Control of Collective Dynamic System (Ph.D. thesis) (Wuhan:Huazhong University of Science and Technolog) (in Chinese)
[13] Shankar S, Souslov A, Bowick M J, Marchetti M C and Vitelli V 2020 arXiv:2010.00364[cond-mat.soft]
[14] Vicsek T, Czirók A, Ben-Jacob E, Cohen I and Shochet O 1995 Phys. Rev. Lett. 75 1226
[15] Meakin P, Vicsek T and Family F 1985 Phys. Rev. B 31 564
[16] Boccaletti S, Latora V, Moreno Y, Chavez M and Hwang D U 2006 Phys. Rep. 424 175
[17] Liu B, Pu Z, Wu S, Shi L, Wang L and Yang W 2022 Improved Self-Propelled Swarms Model with Enhanced Convergence Efficiency:Advances in Guidance, Navigation and Control (Berlin:Springer) pp. 4933-4942
[18] Chen Q S and Ji M 2017 Chin. Phys. B 26 098903
[19] Gro?]mann R, Aranson I S and Peruani F 2020 Nat. Commun. 11 1
[20] Czirók A, Stanley H E and Vicsek T 1997 J. Phys. A:Math. Gen. 30 1375
[21] Zhang H T, Chen M Z and Zhou T 2007 arXiv:0707.3402[physics.data-an]
[22] Hou F, Zhang Y, Fu X, Jiao L and Zheng W 2021 J. Adv. Transport. 2021 9513170
[23] Willard J, Jia X, Xu S, Steinbach M and Kumar V 2020 arXiv:2003.04919[physics.comp-ph]
[24] Zhang Y, Ren J, Wang R, Fang F and Zheng W 2021 Water 13 2095
[25] LeCun Y, Bengio Y. and Hinton G 2015 Nature 521 436
[26] Sarker I H 2021 SN Comput. Sci. 2 1
[27] Mahesh B 2020 Int. J. Sci. Res. 9 381
[28] Chen K and Meng X 2020 J. Comput. Res. Develop. 57 1971 (in Chinese)
[29] Xu R 2018 Application of Machine Learning to Phase Transition Problems (Master thesis) (Xiamen:Xiamen University) (in Chinese)
[30] Z?ttl A 2020 Chin. Phys. B 29 074701
[31] Canabarro A, Fanchini F F, Malvezzi A L, Pereira R and Chaves R 2019 Phys. Rev. B 100 045129
[32] Shchur O, Mumme M, Bojchevski A and Günemann S 2018 arXiv:1811.05868[cs.LG]
[33] Zhang C, Song D, Huang C, Swami A and Chawla N V 2019 Heterogeneous Graph Neural Network:Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining KDD'19 (New York, USA:Association for Computing Machinery) pp. 793-803
[34] Xu Y and Xiaodong Q 2020 Data Analysis and Knowledge Discovery 4 119 (in Chinese)
[35] Deng L, Bing G and Wen Z 2021 International Journal of Web Services Research (IJWSR) 18 63
[36] Chang M B, Ullman T, Torralba A and Tenenbaum J B 2016 arXiv:1612.00341[cs.AI]
[37] Zhang T, Song A and Lan Y 2020 Sci. Sin. Inf. 50 347
[38] Reinhard P G and Suraud E 2008 Introduction to Cluster Dynamics (New York:John Wiley & Sons)
[39] Dulaney A R and Brady J F 2021 Soft Matter 17 6808
[40] Wang R, Fang F, Cui J and Zheng W 2022 Sci. Rep. 12 500
[41] Wang R, Cui J, Zhang Y and Zheng W 2021 J. University of Electronic Science and Technology of China 50 768 (in Chinese)
[42] Battaglia P W, Hamrick J B, Bapst V, et al. 2018 arXiv:1806.01261[cs.LG]
[43] Maron H, Ben-Hamu H, Serviansky H and Lipman Y 2019 arXiv:1905.11136[cs.LG]
[44] Que-Salinas U, Ramírez-Gonzáez P E and Torres-Carbajal A 2021 Soft Matter 17 1975
[45] Dulaney A R and Brady J F 2021 Soft Matter 17 6808
[46] Sanchez-Gonzalez A, Godwin J, Pfaff T, Ying R, Leskovec J and Battaglia P 2020 arXiv:2002.09405[cs.LG]
[47] Zheng W, Zhang S and Xu N 2018 Chin. Phys. B 27 066102
[48] Xu Z, Wang R, Cui J, Liu Y and Zheng W 2021 Chin. Phys. B 30 066101

Graph dynamical networks for forecasting collective behavior of active matter

Graph dynamical networks for forecasting collective behavior of active matter

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 7

Metrics

本文评价

推荐阅读 0

[1]	Huan Liang(梁欢), Peng Liu(刘鹏), Fangfu Ye(叶方富), and Mingcheng Yang(杨明成). Active thermophoresis and diffusiophoresis[J]. 中国物理B, 2022, 31(10): 104702-104702.
[2]	Andreas Zöttl. Simulation of microswimmer hydrodynamics with multiparticle collision dynamics[J]. 中国物理B, 2020, 29(7): 74701-074701.
[3]	李赫, 杨翔, 张何朋. Symmetry properties of fluctuations in an actively driven rotor[J]. 中国物理B, 2020, 29(6): 60502-060502.
[4]	侯章林, 句颖, 宗奕吾, 叶方富, 赵坤. Molecular dynamics simulations on the dynamics of two-dimensional rounded squares[J]. 中国物理B, 2018, 27(8): 88203-088203.
[5]	田波, 孙王平, 李明, 姜锐, 胡茂彬. Flowrate behavior and clustering of self-driven robots in a channel[J]. 中国物理B, 2018, 27(3): 38902-038902.
[6]	陈秋实, 季铭. Collective motion of active particles in environmental noise[J]. 中国物理B, 2017, 26(9): 98903-098903.
[7]	田文得, 顾燕, 郭永坤, 陈康. Anomalous boundary deformation induced by enclosed active particles[J]. 中国物理B, 2017, 26(10): 100502-100502.