Back to search

Clemson University

Graph Based Deep Reinforcement Learning Aided by Transformers for Multi-Agent Cooperation

Abstract

dc:description.abstract

Mission planning for a fleet of cooperative autonomous drones in applications that involve serving distributed target points, such as disaster response, environmental monitoring, and surveil- lance, is challenging, especially under partial observability, limited communication range, and uncertain environments. Traditional path-planning algorithms struggle in these scenarios, particu- larly when prior information is not available. To address these challenges, I propose an innovative framework that integrates Graph Neural Networks (GNNs), Deep Reinforcement Learning (DRL), and transformer-based mechanisms for enhanced multi-agent coordination and collective task ex- ecution. My approach leverages GNNs to model agent-agent and agent-goal interactions through adaptive graph construction, enabling efficient information aggregation and decision-making un- der constrained communication. A transformer-based message-passing mechanism, augmented with edge-feature-enhanced attention, captures complex interaction patterns, while a Double Deep Q-Network (Double DQN) with prioritized experience replay optimizes agent policies in partially observable environments. This integration is carefully designed to address specific requirements of multi-agent navigation, such as scalability, adaptability, and efficient task execution. Experi- mental results demonstrate superior performance, with 90% service provisioning and 100% grid coverage (node discovery), while reducing the average steps per episode to 200, compared to 600 for benchmark methods such as particle swarm optimization (PSO), greedy algorithms and DQN. My framework establishes a robust solution for cooperative Unmanned Aerial Vehicle (UAV) path planning in challenging operational scenarios.

Degree

thesis:*
Name thesis:degree_name
Master of Science (MS)
Level thesis:degree_level
Thesis
Discipline thesis:degree_discipline
Computer Science

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Elrod, Michael S
Contributors dc:contributor
  • Abolfazl Razi
  • Long Cheng
  • Rahul Amin

Subjects

dc:subject × 6

Identifiers

dc:identifier.*
Repository record dc:identifier
https://open.clemson.edu/all_theses/4519

Chain of custody

source
Harvested from
Clemson University
Base URL
open.clemson.edu/do/oai/
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Elrod, Michael S. Graph Based Deep Reinforcement Learning Aided by Transformers for Multi-Agent Cooperation. Thesis thesis, https://open.clemson.edu/all_theses/4519