• Citer

Computed tomography of chemiluminescence using a data-driven sparse sensing framework

par Procacci, Alberto ;Amaduzzi, Ruggero ;Coussement, Axel ;Parente, Alessandro
Référence Applied thermal engineering, 255, 123918
Publication Publié, 2024-10-01

Article révisé par les pairs

• Accès en ligne
• Détails
• Contenu
• Statistiques

Résumé :

The aim of this work is to demonstrate the use of data-driven sparse sensing for the reconstruction of the three-dimensional chemiluminescence field of a flame from the two-dimensional line-of-sight integrated chemiluminescence signal. By leveraging the intrinsic low-dimensionality of physical phenomena, sparse sensing can be employed to reconstruct a signal from few samples. This makes it a good candidate to be employed in computed tomography of chemiluminescence, an imaging technique used to reconstruct the chemiluminescence field from chemiluminescence images. In the proposed methodology, the transforming basis used to project the system's state in the low-dimensional manifold is computed using the proper orthogonal decomposition, which is a popular machine-learning technique for data compression. The methodology is demonstrated on a virtual experiment based on the data coming from the large eddy simulation of a jet flame in a vitiated coflow, where OH is employed as a surrogate of OH∗. This methodology is then compared to conventional techniques that are generally employed to solve tomographic problems. The results show that the proposed technique is able to reconstruct the chemiluminescence field using only one view, compared to the multiple views required by conventional methods. Moreover, our techniques is also capable of predicting the distribution of other features such as temperature and species concentrations.