Date of Award
Master of Science
Department of Operational Sciences
Matthew J. Robbins, PhD
This research formulates and solves the multiagent routing problem with dynamic target arrivals (MRP-DTA), a stochastic system wherein a team of autonomous unmanned aerial vehicles (AUAVs) executes a strike coordination and reconnaissance (SCAR) mission against a notional adversary. Dynamic target arrivals that occur during the mission present the team of AUAVs with a sequential decision-making process which we model via a Markov Decision Process (MDP). To combat the curse of dimensionality, we construct and implement a hybrid approximate dynamic programming (ADP) algorithmic framework that employs a parametric cost function approximation (CFA) which augments a direct lookahead (DLA) model via a parameterization to the objective function. We show a statistically significant improvement over the repeated greedy marginal heuristic benchmark policy for 19 out of 20 problem instances and a statistically significant improvement over the repeated sequential orienteering problem benchmark policy for 8 out of 10 problem instances of the MRP-DTA. Results of excursion analysis show the value trade off of balancing solution quality and computational effort when selecting the base optimization model for our CFA-DLA algorithm.
DTIC Accession Number
Mogan, Andrew E., "Multiagent Routing Problem with Dynamic Target Arrivals Solved via Approximate Dynamic Programming" (2022). Theses and Dissertations. 5350.