다중 란체스터 모형에 대한 실용적 해법

Abstract

We propose a heuristic algorithm for war-game model that is appropriate for warfare in which the maneuver of the attacker is relatively certain. Our model is based on a multi-weapon extention of the Lanchester's square law. However, instead of dealing with the differential equations, we use a multi-period linear approximation which not only facilitates a solution method but also reflects discrete natures of warfare. Then our game model turns out to be a continuous game known to have an -Nash equilibrium for all  ≥. Therefore, our model approximates an optimal warfare strategies for both players as well as an efficient reinforcement of area defense system that guarantees a peaceful equilibrium. Finally, we report the performance of a practical best-response type heuristic for finding an  -Nash equilibrium for a real-scale problem. Keyword:War-Game, Game Theory, -Nash Equilibrium, Heuristic Method