Resilience analysis of multi-state systems with time-dependent behaviors

• A multi-state resilience model is developed based on non-homogeneous semi-Markov reward process. Time-dependent behaviors in both pre- and post-disruption phases are considered. An efficient algorithm for analysis. Most of existing models assume that system performances continuous. In this paper, we consider modeling analysis systems, whose characterized by discrete, rather than continuous variables. Semi-Markov process systems. the model, performance changes, caused either disruptive events or recoveries, modeled as state transitions, rewards used to financial losses incurred during after disruptions. Four metrics defined quantify different aspects resilience. As non-homogeneous, it can capture time-dependent their impact also designed linear interpolation implemented using vectorization . The computational benefits demonstrated through two numerical experiments. We apply method practical case studies, an oil tank farm a re-configurable computing system. results show methods systems accurately efficiently.