Date of Award
Master of Science
Department of Operational Sciences
Andrew J. Geyer, PhD
Aviation fuel is a major component of the Air Force (AF) budget, and vital for the core mission of the AF. This study investigated the viability of LSTMs to increase the accuracy of deterministic NWP models, while also investigating the ability to reduce model generation time. Increased forecast accuracy for wind speeds could be implemented into existing flight path models to further increase fuel efficiency, while reduced modeling times would allow flight planners to generate a flight plan in rapid response situations. The most viable model consisted of an ensemble of six LSTMs trained o six coordinates. The model's error was on average 1.2 m/s higher than the deterministic NWP with a computation time of 1.85 s. The LSTM generated a flight path that was on average 14.2 min slower for an approximately 7 hour 32 min flight. This forecast generation took seconds to complete compared to hours from the deterministic model. While the LSTM architecture in this study was not able to increase forecast accuracy, the speed at which it generates an approximately close forecast can be an integral tool for flight planners in the future.
DTIC Accession Number
Alarcon, Garrett A., "Predicting Upper Atmospheric Weather Conditions Utilizing Long-Short Term Memory Neural Networks for Aircraft Fuel Efficiency" (2020). Theses and Dissertations. 3600.