Loading...
Using Machine Learning to Track the Location of the Shock Train in Hypersonic Engines
Reynolds, Alison
Reynolds, Alison
Abstract
Proposed hypersonic vehicles would be able to travel at five to ten times the speed of sound, but there are still many problems that need to be solved to construct a functioning vehicle. One such problem involves shocks created in the engine isolator when the vehicle reaches high speeds. These shocks must be contained to the isolator to maximize performance and avoid potential failure. This project attempts to track the location of the leading shock given images of airflow from ground tests of engines using random forests and convolutional neural networks. When the models are trained and tested on data from the same facility, the average root mean squared testing error is approximately 18-20mm. However, the models struggle to adapt to the difference in lighting conditions and increase in noise when tested on data from a different facility, leading to much higher error.
Description
Date
2021-05-01
Journal Title
Journal ISSN
Volume Title
Publisher
Collections
Download Dataset
Rights Holder
Usage License
Embargo
Research Projects
Organizational Units
Journal Issue
Keywords
Citation
Department
Mathematics