Constrained system optimization and capability based analysis

In recent years, military analysts have witnessed the increasing use of response surface methods as applied to analysis of large-scale simulation models. Typically, the output of such an analysis is a stochastic response surface. Often this surface is used for subsequent study, perhaps as an objective function in an optimization problem. This paper presents an alternative to the traditional approaches to optimizing a stochastic response surface subject to constraints. It also presents an extension that offers a new way of quickly visualizing complex relationships among variables in a system. First, the paper focuses on the stochastic nature of the response surface and the implications for subsequent optimization and sensitivity analysis. Next, we present a three step process to evaluate and optimize complex systems. Finally, we present a current application: a unique interactive computer model called Capability Based Analysis.