Uncertainty Quantification (UQ) in Complex System of Systems (SoS) Modeling and Simulation (M&S) Environments

Abstract

Prevailing modeling and simulation (M&S) techniques have struggled to provide meaningful quantitative results in M&S of complex system of systems (SoSs) in the face of an environment filled with complex interacting uncertainties. This paper reports on systems thinking applied to “how” M&S techniques should shift to allow a next generation of quantitative tools and techniques. The imperative is to provide quantitative performance results across the constituent interfaces in a modeled architecture. A five-step statistical and parametric algorithm tool that addresses uncertainty quantification (UQ) is presented. [Improving the utility of UQ data evaluation] A quantitative approach to managing complex uncertainties across modeled interfaces using graph theory is proposed. A future vision for SoS engineering (SoSE) that uses graph theory-based modeling is suggested to improve the utility of tools such as UQ is suggested.