Date of Award

3-10-2010

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Electrical and Computer Engineering

First Advisor

Brett Borghetti, PhD

Abstract

Computer Generated Forces (CGF) are used to represent units or individuals in military training and constructive simulation. The use of CGF significantly reduces the time and money required for effective training. For CGF to be effective, they must behave as a human would in the same environment. Real Time Strategy (RTS) games place players in control of a large force whose goal is to defeat the opponent. The military setting of RTS games makes them an excellent platform for the development and testing of CGF. While there has been significant research in RTS agent development, most of the developed agents are only able to exhibit good tactical behavior, lacking the ability to develop and execute overall strategies. By analyzing prior games played by an opposing agent, an RTS agent can determine the opponent's strengths and weaknesses and develop a strategy which neutralizes the strengths and capitalizes on the weaknesses. It can then execute this strategy in an RTS game. This research develops such an RTS agent called the Killer Bee Artificial Intelligence (KBAI). KBAI builds a classifier for an opposing RTS agent which allows it to predict game outcomes. It then takes this classifier, uses it to generate an effective counter-strategy, and executes the tactics required for the strategy. KBAI is both effective and efficient against four high-quality scripted agents: it wins 100% of the time, and it wins quickly. When compared to native artificial intelligence, KBAI has superior performance. It exhibits strategic behavior, as well as the tactics required to execute a developed strategy.

AFIT Designator

AFIT-GCS-ENG-10-07

DTIC Accession Number

ADA516703

Share

COinS