1. Machine learning methods for state-to-state approach.
- Author
-
Campoli, Lorenzo, Fomin, Vasily, and Shiplyuk, Alexander
- Subjects
- *
MACHINE learning , *VISCOUS flow , *EULER equations , *BINARY mixtures , *REGRESSION analysis , *COMPUTER simulation - Abstract
It is well known that numerical simulations of high-speed reacting viscous flows, in the framework of state-to-state (STS) formulations, are often prohibitively computationally expensive. In this work, we investigate the possibility of using machine learning algorithms (MLAs) for STS approaches in order to alleviate such issue. As a first step in this direction, we assessed the potential of data-driven regression models based on machine learning to predict the relaxation terms which appear in the r.h.s. of the Euler system of equations for a one-dimensional reacting shock flow in the STS approach for a N2/N binary mixture. Results show that by appropriately choosing the MLA regressor and opportunely tuning the hyperparameters it is possible to achieve accurate predictions compared to the full-scale STS simulation in significantly shorter times. [ABSTRACT FROM AUTHOR]
- Published
- 2020
- Full Text
- View/download PDF