Artificial intelligence in the modeling of chemical reactions kinetics
-
Maciej Staszak
Abstract
The work presents a selection of recent papers in the field of modeling chemical kinetics by the use of artificial intelligence methods. Due to the fact that kinetics of the chemical reaction is the key element of industrial reactor design and analysis, the work is focused on the presentation of the quality of modeling, the assembly of neural network systems and methods of training required to achieve acceptable results. The work covers a wide range of classes of chemical processes and modeling approaches presented by several authors. Because of the fact that the methods of neural networks training require huge amounts of data, many approaches proposed are intrinsically based on classical kinetics modeling like Monte Carlo methods, quantum ab initio models or classical Arrhenius-like approaches using mass balance rate equations. The work does not fully exhaust the area of artificial intelligence because of its very broad scope and very fast evolution, which has been greatly accelerated recently. However, it is a contribution to describing the current state of science in this field.
-
Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.
-
Research funding: None declared.
-
Conflict of interest statement: The authors declare no conflicts of interest regarding this article.
References
1. Lecun, Y, Bengio, Y, Hinton, G. Deep learning. Nature 2015;521:436–44. Nature Publishing Group. https://doi.org/10.1038/nature14539.Search in Google Scholar PubMed
2. Srivastava, N, Hinton, G, Krizhevsky, A, Salakhutdinov, R. Dropout: a simple way to prevent neural networks from overfitting. J Mach Learn Res 2014;15:1929–58.Search in Google Scholar
3. Palmer, J, Chakravarty, A. Supervised machine learning. In: An introduction to high content screening. Hoboken, NJ, USA: John Wiley & Sons, Inc.; 2015:231–45 pp.10.1002/9781118859391.ch15Search in Google Scholar
4. Oja, E. Finding clusters and components by unsupervised learning. Lect Notes Comput Sci (including Subser Lect Notes Artif Intell Lect Notes Bioinformatics) 2004;3138:1–15. https://doi.org/10.1007/978-3-540-27868-9_1.Search in Google Scholar
5. Feng, Y, Xie, M, Wang, L. Semi-supervised learning method of constructive neural networks. Proceedings of 2015 IEEE advanced information technology, electronic and automation control conference, IAEAC 2015. Chongqing, China: IEEE; 2016:1020–3 p. https://doi.org/10.1109/IAEAC.2015.7428711.Search in Google Scholar
6. Wang, Q, Zhan, Z. Reinforcement learning model, algorithms and its application. Proceedings 2011 international conference on mechatronic science. Electric engineering and computer, MEC 2011. Jilin, China: IEEE; 2011:1143–6 p. https://doi.org/10.1109/MEC.2011.6025669.Search in Google Scholar
7. Irvine, WM. Langmuir–Hinshelwood mechanism. In Encyclopedia of astrobiology. Berlin Heidelberg: Springer; 2011: 905 p.10.1007/978-3-642-11274-4_863Search in Google Scholar
8. Irvine, WM. Eley–Rideal mechanism. In Encyclopedia of astrobiology. Berlin Heidelberg: Springer; 2011:485 p.10.1007/978-3-642-11274-4_502Search in Google Scholar
9. Mora-Briseño, P, Jiménez-García, G, Castillo-Araiza, CO, González-Rodríguez, H, Huirache-Acuña, R, Maya-Yescas, R. Mars van Krevelen mechanism for the selective partial oxidation of ethane. Int J Chem React Eng 2019;17. https://doi.org/10.1515/ijcre-2018-0085.Search in Google Scholar
10. Ross, JRH. The kinetics and mechanisms of catalytic reactions. Contemporary catalysis. Amsterdam, Netherlands: Elsevier; 2019:161–86 p.10.1016/B978-0-444-63474-0.00007-2Search in Google Scholar
11. Microsoft. Learn ML.NET|Free tutorials, courses, videos, and more|.NET. Available from: https://dotnet.microsoft.com/learn/ml-dotnet [Accessed 30 Aug 2020].Search in Google Scholar
12. TensorFlow. Available from: https://www.tensorflow.org/ [Accessed 30 Aug 2020].Search in Google Scholar
13. Keras. Keras: the Python deep learning API. Available from: https://keras.io/ [Accessed 30 Aug 2020].Search in Google Scholar
14. ONNX. ONNX|Home. Available from: https://onnx.ai/ [Accessed 30 Aug 2020].Search in Google Scholar
15. Barwey, S, Raman, V. A neural network inspired formulation of chemical kinetics [Online]; 2020. Available from: http://arxiv.org/abs/2008.08483 [Accessed 02 Sep 2020].Search in Google Scholar
16. Goodwin, DG, Speth, RL, Moffat, HK, Weber, BW. Cantera: an object-oriented software toolkit for chemical kinetics, thermodynamics, and transport processes. Zenodo; 2018. https://doi.org/10.5281/ZENODO.1174508.Search in Google Scholar
17. Yang, F, Dai, C, Tang, J, Xuan, J, Cao, J. A hybrid deep learning and mechanistic kinetics model for the prediction of fluid catalytic cracking performance. Chem Eng Res Des 2020;155:202–10. https://doi.org/10.1016/j.cherd.2020.01.013.Search in Google Scholar
18. Kollenz, P, Herten, DP, Buckup, T. Unravelling the kinetic model of photochemical reactions via deep learning. J Phys Chem B 2020;124:6358–68. https://doi.org/10.1021/acs.jpcb.0c04299.Search in Google Scholar PubMed
19. Ansys. Chemkin-Pro: chemistry effects predicting simulation software|Ansys. Available from: https://www.ansys.com/products/fluids/ansys-chemkin-pro [Accessed 29 Aug 2020].Search in Google Scholar
20. Baş, D, Dudak, FC, Boyaci, IH. Modeling and optimization III: reaction rate estimation using artificial neural network (ANN) without a kinetic model. J Food Eng 2007;79:622–8. https://doi.org/10.1016/j.jfoodeng.2006.02.021.Search in Google Scholar
21. Grambow, CA, Pattanaik, L, Green, WH. Deep learning of activation energies. J Phys Chem Lett 2020;11:2992–7. https://doi.org/10.1021/acs.jpclett.0c00500.Search in Google Scholar PubMed PubMed Central
22. Swanson, K. Message passing neural networks for molecular property prediction. Massachusetts, USA: Massachusetts Institute of Technology; 2019.Search in Google Scholar
23. Van Der Maaten, L, Hinton, G. Visualizing data using t-SNE. J Mach Learn Res 2008;9:2579−605.Search in Google Scholar
24. Bracconi, M, Maestri, M. Training set design for machine learning techniques applied to the approximation of computationally intensive first-principles kinetic models. Chem Eng J 2020;400:125469. https://doi.org/10.1016/j.cej.2020.125469.Search in Google Scholar
25. Reuter, K, Scheffler, M. Composition, structure, and stability of (formula presented) as a function of oxygen pressure. Phys Rev B Condens Matter 2002;65:1–11. https://doi.org/10.1103/PhysRevB.65.035406.Search in Google Scholar
26. Sharma, AJ, Johnson, RF, Kessler, DA, Moses, A. Deep learning for scalable chemical kinetics. AIAA scitech 2020 forum. Orlando, USA: American Institute of Aeronautics and Astronautics (AIAA); 2020. https://doi.org/10.2514/6.2020-0181.Search in Google Scholar
27. Melkumova, LE, Shatskikh, SY. Comparing ridge and LASSO estimators for data analysis. Procedia Eng 2017;201:746–55. https://doi.org/10.1016/j.proeng.2017.09.615.Search in Google Scholar
28. Marcuś, M, Conaire, M, Curran, HJ, Simmie, JM, Pitz, WJ, Westbrook, CK. A comprehensive modeling study of hydrogen oxidation; 2015.Search in Google Scholar
29. Mills, K, Spanner, M, Tamblyn, I. Deep learning and the Schrödinger equation. Phys Rev A 2017;96:042113. https://doi.org/10.1103/PhysRevA.96.042113.Search in Google Scholar
30. Truong, TN, Truhlar, DG. Ab initio transition state theory calculations of the reaction rate for OH + CH4 → H2O+CH3. J Chem Phys 1990;93:1761–9. https://doi.org/10.1063/1.459103.Search in Google Scholar
© 2020 Walter de Gruyter GmbH, Berlin/Boston
Articles in the same Issue
- Frontmatter
- Reviews
- Multiscale modeling and simulation of magneto-active elastomers based on experimental data
- Theoretical examination of efficiency of anthocyanidins as sensitizers in dye-sensitized solar cells
- Artificial intelligence in the modeling of chemical reactions kinetics
- Computational studies of biologically active alkaloids of plant origin: an overview
- Certainty through uncertainty: stochastic optimization of grid-integrated large-scale energy storage in Germany
- Shaping the future energy markets with hybrid multimicrogrids by sequential least squares programming
Articles in the same Issue
- Frontmatter
- Reviews
- Multiscale modeling and simulation of magneto-active elastomers based on experimental data
- Theoretical examination of efficiency of anthocyanidins as sensitizers in dye-sensitized solar cells
- Artificial intelligence in the modeling of chemical reactions kinetics
- Computational studies of biologically active alkaloids of plant origin: an overview
- Certainty through uncertainty: stochastic optimization of grid-integrated large-scale energy storage in Germany
- Shaping the future energy markets with hybrid multimicrogrids by sequential least squares programming
