Linear programming based models for resilient supply systems design

Abstract

This thesis dedicates to develop a quantitative framework of modeling the system resilience which can guide the design of general supply systems under an uncertain environment. In particular, we propose a formal definition of resilience measure by analyzing the penalty caused by demand loss against a tolerance level. The precise definition of resilience measure is given by a number of self-contained axioms that reflect some common decision orientations along with computational considerations. After that, we study some interesting perspectives of resilience measure and apply our framework to two distinct applications. The results suggest the superiority of our proposed resilience measure based framework in providing good stochastic system performance under uncertainty over some traditional methods.