Abstract
This paper considers an optimization-based approach to battle management. The framework consists of two opposing forces of various resources subject to attrition and evolving over an arena of sectors. The resulting model is a large-scale linear flow subject to various positivity constraints. It can be shown that when resource mobility constraints and attrition losses are taken into account, the problem of finding optimal tactics is computationally intractable. We pursue an alternative approach that allows the use of linear programming in conjunction with receding horizon control. The main idea is to derive model simplifications that alleviate the computational burden while providing suitable tactics. (Author)