Advisor(s)
Roy Koomullil
Committee Member(s)
David Littlefield
Leon Jololian
School
School of Engineering
Document Type
Thesis
Department (new version)
Engineering
Date of Award
9-9-2024
Degree Name by School
Master of Mechanical Engineering (MME) School of Engineering
Abstract
Predicting wind and temperature fields around military vehicles during extended missions is crucial to avoid detectability by infrared (IR) devices. Also, the abrupt shifting of wind direction can have a significant impact on vehicle stability. This is a challenging task due to the vehicles' geometric complexity and the unpredictable nature of wind direction, which can shift abruptly. Computational Fluid Dynamics (CFD) is routinely used for calculating the flow fields around ground vehicles. However, this requires extensive computational time and memory, making it unsuitable for real-time analysis. To address these challenges, this research focuses on machine learning (ML) techniques for accurate wind field prediction in real-time for any arbitrary wind direction. Reduced Order Modeling (ROM) is used for the dimensionality reduction of flow field data derived from high-fidelity CFD simulations. ML models are trained using the low-dimensional data from ROM, and the predicted low-dimensional data for any given wind direction by the trained ML model is used to reconstruct the flow field. ROM, in conjunction with ML techniques, offers a substantial reduction in analysis time while maintaining the ability to predict the flow field accurately. In this study, three different neural network architectures were used for the predictions, and their accuracy was evaluated by comparing them with the CFD results. The optimal ML model is identified by varying the number of hidden layers and neurons within those layers.
ProQuest ID
Recommended Citation
Ramogi, Emmanuel Ong'aro, "Reduced Order Modeling (Rom) Using Machine Learning Techniques For Analyzing Fluid Flow Around A Ground Vehicle" (2024). All ETDs from UAB. 7635.
https://digitalcommons.library.uab.edu/etd-collection/7635
