Abstract
The parameters of a level-set flame model are inferred using an ensemble of heteroscedastic Bayesian neural networks (BayNNEs). The neural networks are trained on a library of 1.7 million observations of 8500 simulations of the flame edge, obtained using the model with known parameters. The ensemble produces samples from the posterior probability distribution of the parameters, conditioned on the observations, as well as estimates of the uncertainties in the parameters. The predicted parameters and uncertainties are compared to those inferred using an ensemble Kalman filter. The expected parameter values inferred with the BayNNE method, once trained, match those inferred with the Kalman filter but require less than one millionth of the time and computational cost of the Kalman filter. This method enables a physics-based model to be tuned from experimental images in real time.
M. L. Croci and U. Sengupta—Equal contribution.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Cantera is a suite of tools for problems involving chemical kinetics, thermodynamics, and transport processes [13].
- 2.
The EnKF is fully parallelised: the processors have 32 cores in total, one for each member in the ensemble.
References
Juniper, M.P., Sujith, R.: Sensitivity and nonlinearity of thermoacoustic oscillations. Ann. Rev. Fluid Mech. 50, 661–689 (2018)
Keller, J.J.: Thermoacoustic oscillations in combustion chambers of gas turbines. AIAA 33–12 (1995)
Strutt, J.W.: The Theory of Sound, vol. II. Macmillan and Co., London (1878)
Smart, A.E., Jones, B., Jewel, N.T.: Measurements of unsteady parameters in a rig designed to study reheat combustion instabilities. AIAA 26-88 (1976)
Pitz, R.W., Daily, J.W.: Experimental study of combustion in a turbulent free shear layer formed at a rearward facing step. AIAA 81-106 (1981)
Smith, D.A., Zukoski, E.E.: Combustion instability sustained by unsteady vortex combustion. AIAA 85-1248 (1985)
Poinsot, T., Trouve, A., Veynante, D., Candel, S., Esposito, E.: Vortex-driven acoustically coupled combustion instabilities. J. Fluid Mech. 177–220, 265–292 (1987)
Crocco, L.: Research on combustion instability in liquid propellant rockets. In: Symposium (International) Combustion, vol. 12, no. 1, pp. 85–99 (1969)
Williams, F.A.: Turbulent combustion. In: The Mathematics of Combustion, pp. 97–131 (1985)
Evensen, G.: Sequential data assimilation with a nonlinear quasi-geostrophic model using Monte Carlo methods to forecast error statistics. J. Geophys. Res. 99, 10143–10162 (1994)
Yu, H., Juniper, M.P., Magri, L.: Combined state and parameter estimation in level-set methods. J. Comput. Phys. 399, 108950 (2019)
Yu, H., Juniper, M.P., Magri, L.: A data-driven kinematic model of a ducted premixed flame. Proc. Combust. Inst. 399, 6231–6239 (2020)
Goodwin, D.G., Speth, R.L., Moffat, H.K., Weber, B.W.: Cantera: an object-oriented software toolkit for chemical kinetics, thermodynamics, and transport processes (2018). https://www.cantera.org. Version 2.4.0
Gal, Y.: Uncertainty in deep learning. Ph.D. thesis (2016)
Damianou, A.C., Lawrence, N.D.: Deep Gaussian processes. In: Proceedings of the 16th International Conference on Artificial Intelligence and Statistics (AISTATS) (2013)
MacKay, D.J.C.: Information Theory, Inference and Learning Algorithms. Cambridge University Press, Cambridge (2003)
Gal, Y., Ghahramani, Z.: Dropout as a Bayesian approximation: representing model uncertainty in deep learning. In: Proceedings of the 33rd International Conference on Machine Learning. NY, USA, New York (2016)
Pearce, T., Zaki, M., Brintrup, A., Anastassacos, N., Neely, A.: Uncertainty in neural networks: Bayesian ensembling. In: International Conference on Artificial Intelligence and Statistics (AISTATS) (2020)
Sengupta, U., Croci, M.L., Juniper, M.P.: Real-time parameter inference in reduced-order flame models with heteroscedastic Bayesian neural network ensembles. In: Machine Learning and the Physical Sciences Workshop at the 34th Conference on Neural Information Processing Systems (NeurIPS) (2020)
Sengupta, U., Amos, M., Hosking, J.S., Rasmussen, C.E., Juniper, M.P, Young, P.J.: Ensembling geophysical models with Bayesian neural networks. In: Advances in Neural Information Processing Systems (NeurIPS) (2020)
Kashinath, K., Li, L.K.B., Juniper, M.P.: Forced synchronization of periodic and aperiodic thermoacoustic oscillations: lock-in, bifurcations and open-loop control. J. Fluid Mech. 838, 690–714 (2018)
Hemchandra, S.: Dynamics of turbulent premixed flames in acoustic fields. Ph.D. thesis (2009)
Lakshminarayanan, B., Pritzel, A., Blundell, C.: Simple and scalable predictive uncertainty estimation using deep ensembles. In: Advances in Neural Information Processing Systems (NeurIPS), pp. 6402–6413 (2017)
Luo, X., Hoteit, I.: Robust ensemble filtering and its relation to covariance inflation in the ensemble Kalman filter. Monthly Weather Rev. 139(12), 3938–3953 (2011)
Funding
This project has received funding from the UK Engineering and Physical Sciences Research Council (EPSRC) award EP/N509620/1 and from the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement number 766264.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
A Supplementary material: Hyperparameter settings
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Croci, M.L., Sengupta, U., Juniper, M.P. (2021). Data Assimilation Using Heteroscedastic Bayesian Neural Network Ensembles for Reduced-Order Flame Models. In: Paszynski, M., Kranzlmüller, D., Krzhizhanovskaya, V.V., Dongarra, J.J., Sloot, P.M. (eds) Computational Science – ICCS 2021. ICCS 2021. Lecture Notes in Computer Science(), vol 12746. Springer, Cham. https://doi.org/10.1007/978-3-030-77977-1_33
Download citation
DOI: https://doi.org/10.1007/978-3-030-77977-1_33
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-77976-4
Online ISBN: 978-3-030-77977-1
eBook Packages: Computer ScienceComputer Science (R0)