Methods and models in preparing weapon-target interaction data for combat simulations

Abstract

Combat simulations are part of a suite of tools used by Defence Science and Technology Organisation (DSTO) to support Army decision makers, allowing analysis of the effects of changes in equipment, tactics or force structure. In order to represent combat effectively, these tools require enormous amounts of input data, which cannot be gained from empirical sources. This input data must cover aspects such as detection and identification of targets, behavioral decision making, basic system capabilities and weapon/target interactions. This data is represented within these combat simulations as lookup tables, from which an appropriate result, or probability of a result, is selected. Our approach (Mazonka, 2012) seeks to provide fit-for-purpose vulnerability and lethality data to allow these simulations to adjudicate the outcomes of combat. This approach takes simple physical data from weapon and target systems and applies physics models to determine the probabilistic results of their interactions. We are cognisant of a number of limitations: the paucity of available empirical data, short lead times for data requirements and a requirement for an extremely broad set of interaction data, at the expense of depth. In this paper we briefly describe our solutions to problems of armour penetration, probability propagation and blast effects, along with methods for converting generic result data such as vehicle probability of kill models into simulation-specific input data. These solutions are presented as a series of limited, low level methods, which illustrate some of the challenges inherent in generating fit-for-purpose combat simulation data.