Adaptive battle agents: complex adaptive combat models
Kim L. Lim, Isaac Mann, Ricardo Santos, Brad Tobin, Matthew J. Berryman, Derek Abbott, Alex Ryan · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2005
This work explores emergent behaviour in a complex adaptive system, specifically an agent-based battlefield simulation model. We explore the changes in agent attribute sets through the use of genetic algorithms over a series of battles, with performance measured by a number of different statistics including number of casualties, number of enemy agents killed, and success rate at "capturing the flag". The agents' capabilities include (but are not limited to) maneuverability upon the battlefield, formulating, sending, receiving and acting upon messages and attacking enemy agents.