Date of Graduation

12-2022

Document Type

Thesis

Degree Name

Master of Science in Industrial Engineering (MSIE)

Degree Level

Graduate

Department

Industrial Engineering

Advisor/Mentor

Pohl, Edward A.

Committee Member

Liao, Haitao

Second Committee Member

Sullivan, Kelly M.

Keywords

Machine learning; Network reliability; Reinforcement learning; Resource allocation

Abstract

Networks provide a variety of critical services to society (e.g. power grid, telecommunication, water, transportation) but are prone to disruption. With this motivation, we study a sequential decision problem in which an initial network is improved over time (e.g., by adding or increasing the reliability of edges) and rewards are gained over time as a function of the network’s all-terminal reliability. The actions during each time period are limited due to availability of resources such as time, money, or labor. To solve this problem, we utilized a Deep Reinforcement Learning (DRL) approach implemented within OpenAI-Gym using Stable Baselines. A Proximal Policy Optimization (PPO) was used to identify the edge to be improved or a new edge to be added based on the current state of the network and the available budget. To calculate the network’s all-terminal reliability, a reliability polynomial was employed. To understand how the model behaves under a variety of conditions, we explored numerous network configurations with different initial link reliability, added link reliability, number of nodes, and budget structures. We conclude with a discussion of insights gained from our set of designed experiments.

Share

COinS