Date of Graduation

7-2026

Document Type

Thesis

Degree Name

Master of Science in Industrial Engineering (MSIE)

Degree Level

Graduate

Department

Industrial Engineering

Advisor/Mentor

Curry, Rob

Committee Member

Shen, Haoming

Second Committee Member

Sullivan, Kelly

Keywords

High-Value Unit (HVU) Survivability; Integrated Simulated Annealing with Dynamic Programming (SA-DP); Maritime Chokepoints; Piecewise McCormick Relaxation; Two-stage Stochastic Optimization

Abstract

Protecting high-value units (HVUs) while traversing maritime chokepoints under threats is a significant challenge that requires effective planning for allocating limited naval assets (escort ships, detection capabilities, and defensive capabilities). Although the weapon-to-target assignment problem has been previously studied in the literature, the problem of allocating limited naval assets under uncertain threat scenarios remains unexplored. This research develops a two-stage stochastic optimization framework with the objective of maximizing the expected weighted survivability of HVUs across all scenarios. The first-stage decisions include assigning available escorts to HVUs, selecting locations for deploying detection capabilities, and allocating defensive capabilities to each escort prior to threat uncertainty. The second-stage decisions include which detection capabilities detect each threat, and which set of defensive capabilities engage each threat under the revealed uncertainty. We develop three unique solution approaches for solving this problem: mixed-integer nonlinear programming formulated in GEKKO, linear approximation using piecewise McCormick relaxation, and an integrated simulated annealing with dynamic programming (SA-DP) framework. Computational experiments on five unique problem instances suggest that the proposed SA-DP approach produces better solutions in less computational time than the other two approaches in four of the five instances. This framework can be extended to relay-based escort operations, which would allow escorts to be reassigned among multiple HVUs across different stages of the transit.

Download

Share

COinS