Display Full Text | Download Full Text
1.49 MB PDF file

Degree

Master of Science in Engineering (MSEgr), Wright State University, Human Factors Engineering, 2007.

Abstract

In many current agent-based modeling systems, it is difficult for a domain-expert user to define and implement agent behaviors without possessing extensive programming knowledge. MUAVES is an existing simulation environment that serves as a research testbed for examining command and control issues with unmanned aerial vehicle (UAV) systems containing many vehicle agents. In its previous form, defining agent behaviors required knowledge of the C# programming language that some MUAVES users did not have. This thesis presents a new methodology for the definition and implementation of UAV agent behaviors in MUAVES. The new methodology is based on diagramming an agent’s controller state. No programming knowledge is required to reuse modular behaviors and trigger conditions specified by previous researchers. The definition of novel behaviors has also been improved by placing behavioral code in external library files, away from the main simulation code. These novel behaviors can be implemented at any desired level of abstraction. After describing the methodology, some sample scenarios are presented as proofs-of-concept.

Keywords

agent; behavior; implementation; definition; unmanned aerial vehicle; uav; methodology; muaves; abstraction; modeling

Advisor

Raymond R. Hill

Pages

90p.

Document number: wright1175104194
Permalink: http://rave.ohiolink.edu/etdc/view?acc_num=wright1175104194

This ETD has been downloaded 551 times (through March 2013)