Theses and Dissertations
Issuing Body
Mississippi State University
Advisor
Peterson, Luke
Committee Member
To, Filip Sumunto D.
Committee Member
Lee, Nayeon
Committee Member
Pezzola, Genevieve
Committee Member
Elder, Steven H.
Date of Degree
12-9-2022
Document Type
Graduate Thesis - Open Access
Major
Biomedical Engineering
Degree Name
Master of Science (M.S.)
College
James Worth Bagley College of Engineering
Department
Department of Agricultural and Biological Engineering
Abstract
Artificial Neural Network (ANN) ensemble and Response Surface Method (RSM) surrogate models were generated from Finite Element (FE) simulations to predict the overpressure load and vehicle impact response of a novel rapidly deployable protective structure. A Non-dominated Sorting Genetic Algorithm-II (NSGA-II) was used in conjunction with the surrogate models to determine structure topology input variable configurations which were suited to produce the optimal balance of minimum mass, minimum rotation angle, minimum displacement, and maximum total length of the deployable structure. The structure was designed to retract into a container, be lightweight to facilitate transportation, and be able to adapt to varying terrain slopes. This research demonstrates that, in comparison to the RSM, ANN ensembles can more accurately and efficiently be used for identifying optimal design solutions for multi-objective design problems when two surrogate models from the same method corresponding to separate FE models are used simultaneously in a NSGA-II.
Recommended Citation
Tellkamp, Daniela F., "Surrogate model-based design optimization of a mobile deployable structure for overpressure load and vehicular impact mitigation" (2022). Theses and Dissertations. 5671.
https://scholarsjunction.msstate.edu/td/5671