Exploring Trade-offs and Emergent Properties of Heterogeneous Swarms of Maritime Robot Systems through Empirical Analysis and Application-Driven Experiments

Multi-agent systems present a promising approach to addressing challenges such as searching for and tracking moving targets, offering advantages like robustness and scalability over single-agent solutions. Current maritime searching and tracking strategies typically involve employing predetermined paths for exploring the search space or adopting Particle Swarm Optimization (PSO) algorithms for multi-robot systems (MRS). While these approaches often entail homogeneous deployment of algorithms or behaviors across all agents, the potential benefits of introducing heterogeneity still need to be explored. Specifically, varying agent behavior or capabilities could enhance mission performance, but the trade-offs involved are not thoroughly understood beyond trivial outcomes like adjusting agent speed. In this thesis, a novel swarming approach is used to tackle two core missions: dynamic target searching and tracking, and isocontour identification. The strategy employs a combination of five distinct algorithms. The innovation lies in introducing heterogeneity among agents by assigning specific roles managed through varying weights tied to each algorithm. Trade-offs between mission performance and cost are quantified by simulating a swarm with diverse roles and behaviors. Key performance metrics include the accuracy of target position estimation, convergence time on target, duration of target tracking, and correlation between the swarm's collective heading and target bearing. The overall energy consumption of the swarm determines the cost metric. Investigation of the impact due to different proportions of agent types within the swarm, provided valuable insights into how to optimize mission effectiveness while managing resource constraints. Overall, this research contributes to advancing the understanding of how heterogeneity in multi-agent systems can enhance mission performance and offers practical insights into optimizing resource allocation in complex tasks such as target search and tracking. By comprehensively assessing trade-offs, this thesis aims to pave the way for more efficient and adaptable multi-agent systems in real-world applications.

Name thesis:degree_name

Master

Department dc:contributor.department

Massachusetts Institute of Technology. Department of Mechanical Engineering

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2024

Author dc:creator

• Hoang, Thinh B.

Advisor dc:contributor.advisor

• Yue, Dick K. P.

Statement dc:rights

• In Copyright - Educational Use Permitted
• Copyright retained by author(s)

Licence dc:rights.uri

• https://rightsstatements.org/page/InC-EDU/1.0/

Handle dc:identifier.uri

https://hdl.handle.net/1721.1/155655

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/155655