Computational Technology Resources - CCC

Z. Sun¹, H. Du¹, C. Miao², and Q. Hou²

¹College of Intelligence and Computing, Tianjin University, China
²School of Civil Engineering, Tianjin University, China

doi:10.4203/ccc.2.6.3

Z. Sun, H. Du, C. Miao,, Q. Hou, "PINN-ADR: A neural network based simulation tool for reacting flow with multicomponent reactants", in B.H.V. Topping, P. Iványi, (Editors), "Proceedings of the Eleventh International Conference on Engineering Computational Technology", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 2, Paper 6.3, 2022, doi:10.4203/ccc.2.6.3

Keywords: physics-informed neural networks, inverse problem, reactive flow, advection-diffusion-reaction equation, deep learning, autocatalytic reaction.

Abstract

The aim of this work is to analyse the use of PINN to solve forward and inverse problems of reacting flow with multicomponent reactants. For the above two problems, PINN can successfully get the correct result. In the forward problem, using the sin function as the activation function fits the discontinuous boundary problem better than using tanh, and the training speed is faster. In the inverse problem, PINN uses data to learn model parameters can not only learn the convection term and the diffusion coefficients, but also learn the parameters related to the kinetics of the reaction and mutation efficiency, and heuristically guide us to discover the physical and chemical laws in the reaction flow.

download the full-text of this paper (PDF, 9 pages, 704 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description

	Computational & Technology Resources an online resource for computational, engineering & technology publications
	not logged in - login
Front Page Browse CCC CCP CSETS CTR IJRT Other Authors Search Purchase Guide FAQ Contact us	Civil-Comp Conferences ISSN 2753-3239 CCC: 2 PROCEEDINGS OF THE ELEVENTH INTERNATIONAL CONFERENCE ON ENGINEERING COMPUTATIONAL TECHNOLOGY Edited by: B.H.V. Topping and P. Iványi Paper 6.3 PINN-ADR: A neural network based simulation tool for reacting flow with multicomponent reactants Z. Sun¹, H. Du¹, C. Miao², and Q. Hou² ¹College of Intelligence and Computing, Tianjin University, China ²School of Civil Engineering, Tianjin University, China doi:10.4203/ccc.2.6.3 Full Bibliographic Reference for this paper Z. Sun, H. Du, C. Miao,, Q. Hou, "PINN-ADR: A neural network based simulation tool for reacting flow with multicomponent reactants", in B.H.V. Topping, P. Iványi, (Editors), "Proceedings of the Eleventh International Conference on Engineering Computational Technology", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 2, Paper 6.3, 2022, doi:10.4203/ccc.2.6.3 Keywords: physics-informed neural networks, inverse problem, reactive flow, advection-diffusion-reaction equation, deep learning, autocatalytic reaction. Abstract The aim of this work is to analyse the use of PINN to solve forward and inverse problems of reacting flow with multicomponent reactants. For the above two problems, PINN can successfully get the correct result. In the forward problem, using the sin function as the activation function fits the discontinuous boundary problem better than using tanh, and the training speed is faster. In the inverse problem, PINN uses data to learn model parameters can not only learn the convection term and the diffusion coefficients, but also learn the parameters related to the kinetics of the reaction and mutation efficiency, and heuristically guide us to discover the physical and chemical laws in the reaction flow. download the full-text of this paper (PDF, 9 pages, 704 Kb) go to the previous paper go to the next paper return to the table of contents return to the volume description
Back to top	©Civil-Comp Limited 2023 - terms & conditions