A generative adversarial network (GAN) approach to creating synthetic flame images from experimental data
Anthony Carreon, Shivam Barwey, Venkat Raman
Abstract
Modern diagnostic tools in turbulent combustion allow for highly-resolved measurements of reacting flows; however, they tend to generate massive data-sets which may cause conventional analysis to be intractable and inefficient. To alleviate this problem, machine learning tools may be used to, for example, discover features from the data for downstream modeling and prediction tasks. To this end, this work applies generative adversarial networks (GANs) to generate realistic flame images based on a time-resolved data set of hydroxide concentration snapshots obtained from planar laser induced fluorescence measurements of a model combustor. The trained networks are able to generate flames in attached, lifted, and intermediate configurations. Using k-means clustering and proper orthogonal decomposition, the synthetic image set produced by the GAN is shown to be visually similar to the real image set, with recirculation zones and burned/unburned regions clearly present. To the authors’ knowledge, this work is the first to demonstrate GAN usage for purely generative modeling in turbulent combustion. Combined with techniques for controlling the configuration of generated flames, new avenues are provided for understanding and predicting flame instability, transition, and flashback.