Risk-Based Policy Optimization for Critical Infrastructure Resilience against a Pandemic Influenza Outbreak

Abstract

Decisions regarding infrastructure resilience are made under uncertainty and involve trade-offs among competing objectives. An effective way of understanding how uncertainties propagate and understanding trade-offs among multiple objectives is to use a computer simulation that integrates high-level representations of each infrastructure, their interdependencies, and their reactions to a variety of potential disruptions. To address this need for such a decision support tool, this paper considers a multidisciplinary federation of systems dynamics models for the purpose of simulating the response of critical and interconnected homeland infrastructures to a major disruption. With the addition of disease progression simulation, the models provide a high-level integrated analysis of pandemic influenza outbreak. By use of the models, options for mitigation and prevention such as the use of antivirals, surgical masks, and quarantine policies can be assessed. In this paper, the models are augmented with analytical methods of uncertainty analysis to determine the statistics of model outputs from the statistics of model inputs in order to determine the relative importance of the uncertainties in the model inputs and to identify the worst-case scenarios that have a given probability of occurrence. Techniques of reliability-based optimization are incorporated to find a robust optimal strategy for infrastructure resilience in which preparations are optimized for the worst-case scenario.