Comparing combat models using analytical surrogates.

• Main Author: Green, John R.
• Other Authors: Barr, Donald R.
• Language: en_US
• Published: Monterey, California. Naval Postgraduate School 2013
• Online Access: http://hdl.handle.net/10945/26415; http://hdl.handle.net/10945/26415

The widespread availability of inexpensive high-speed computers has led to the development of complex, detailed technical models of combat. These high resolution computer simulations and wargames are touted by their proponents as low-cost alternatives to extensive, high-cost field training exercises for the training of combat leaders. The validity of these simulations as models of combat, and thus as useful training tools is unproven. Direct comparison of simulations with field training exercises is often frustrated by the inherent complexities in each, and the shortage of quality data from field exercises. This thesis examines the feasibility of comparing these systems indirectly through the use of surrogate analytical models. A simple discrete time stochastic surrogate model is examined. Techniques for using the surrogate model to compare battle data are studied using simulated data from a simple combat model. Areas for further research are discussed. Combat models, Simulated annealing, Regression, Difference equations, Stochastic models