Date of Award

4-2017

Document Type

Thesis - Open Access

Degree Name

Master of Science in Mechanical Engineering

Department

Mechanical Engineering

Committee Chair

Eric J. Coyle, Ph.D.

First Committee Member

Patrick N. Currier, Ph.D.

Second Committee Member

Marc D. Compere, Ph.D.

Abstract

A situation is addressed in which multiple autonomous agents are used to search an unknown environment for a target, the position and orientation of which is known with respect to each agent. A controlling framework is proposed to inform and coordinate the agents’ movements in order to reduce the time required to locate the target. Four primary variables are considered: the cost function used to select the agents’ paths, the number of agents in a given scenario, the distance over which the agents are assumed to communicate, and the size of the environment in which the agents are operating. It was found that a cost function that balances progress toward the target with exploration of the environment is generally most effective for all combinations of the other variables. More agents and greater communication are beneficial, to a point, in larger environments, although these may be less effective in smaller ones.

Share

COinS