Path Planning and Autonomous Navigation for use in Computer Generated Forces

In this Master´s thesis project, the use of Unmanned Combat Aerial Vehicle, UCAV, in the fighter role was studied. The mission of the agent was to guard and supervise a No-Fly Zone (NFZ) and to, upon violation, intercept a single intruder while minimizing the risk of being shot down. Methods for on-line path planning on a discretized version of the search space were evaluated and, in order to further enhance paths, post processing using virtual forces was applied. Simulations showed that threat-constrained path planning, based on a tradeoff between path length and risk estimate, can be performed in satisfyingly short time, suitable for on-line use. Further considerations have yet to be made in order to meet the demands of an uncertain environment with dynamic threats and targets maneuvering in order to achieve diversion. As part of the project, a prototype of the agent was implemented as an add-on to a fighter aircraft entity running in the Mission Training simulator at the Swedish Air Force Combat Simulation Centre (FLSC). It is finally suggested that path planning procedures, like the ones studied in the project, should be an integrated part in future tools for Computer Generated Forces (CGF) as they could cover some of the shortcomings in pure rule-based systems.