Keywords and Phrases
"First responder training is becoming ever more important in our world. Several techniques are currently being explored to train first responders in physical and virtual worlds. The First Responder Simulation and Training Environment is being developed to augment the physical training efforts. This environment has many different components including a gas model, virtual environment, sensor simulator, and human interface equipment. An addition being made to the system is autonomous agents to interact with the first responders. Genetic programming, an evolutionary computation technique, is being used to evolve agents to interact in the environment. Friendly, hostile and neutral agents are being evolved in a two dimensional grid environment. Expression and decision trees are used to represent the individuals being evolved. The terminals for the decision trees are analyzed for their effectiveness. Low selective pressure and high crossover versus mutation rate were found to produce good individuals. Genetic programming is successfully employed to generate high performance individuals in the environment"--Abstract, page iv.
Hilgers, Michael Gene
Tauritz, Daniel R.
Hall, Richard H.
M.S. in Computer Science
University of Missouri--Rolla
ix, 32 pages
© 2004 Alex Jerome Berry, All rights reserved.
Thesis - Restricted Access
Intelligent agents (Computer software)
Virtual reality -- Technological innovations
Print OCLC #
Link to Catalog Record
Electronic access to the full-text of this document is restricted to Missouri S&T users. Otherwise, request this publication directly from Missouri S&T Library or contact your local library.http://laurel.lso.missouri.edu/record=b5406675~S5
Berry, Alex J., "Evolving intelligent agents for adaptive first responder virtual training" (2004). Masters Theses. 3644.
Share My Thesis If you are the author of this work and would like to grant permission to make it openly accessible to all, please click the button above.