Real-time parameter inference of nonlinear bluff-body-stabilized flame models using Bayesian neural network ensembles

This study uses a Bayesian machine learning method to infer the parameters of a physics-based model of a bluffbody-stabilized premixed flame in real-time. An ensemble of neural networks is trained on a library of simulated flame fronts with known parameters, generated using a level-set solver, LSGEN2D. This trained ensemble then observes experimental images of a qualitatively similar flame. The ensemble provides reliable estimates of the parameters and their uncertainties, from which the flame can be re-simulated beyond the observation window of the experiment. Using the re-simulated flame, the flame surface area, a proxy for the heat release rate, is calculated. The method is general: once trained, the ensemble can be used to infer the parameters from any bluff-body-stabilized premixed flame as long as the flame is qualitatively similar and the parameters lie within the ranges in the training library. Recognizing each set of 10 frames takes milliseconds, which is fast enough to work in real-time.