Using a competitive approach to improve military simulation artificial intelligence design

Abstract

The research presented in this thesis attempts to show how using a competitive approach to artificial intelligence (AI) design can lead to improvement of the AI solutions used in military simulations. To demonstrate the potential of the competitive approach, ORTS, a real-time strategy game engine, and its competition setup are used. To justify the thesis statement, a set of exploratory experiments are conducted. The experiments represent a tournament of virtual battles between base case AIs and test case AIs. The existing OTRS clients are used as base cases, and the test cases are evolved using the competitive approach to AI design described in this work. The analysis of the results from the tournament proves the advantages of the competitive approach. At the end of the thesis, some conclusions and recommendations for future work are made.