Time Granularity in System-of-Systems Simulation of Infrastructure Networks

摘要

Because of their extreme complexities, a system-of-systems (SoS) approach is often used for simulating infrastructure systems. This allows the user to integrate models of various systems into one simulation. However, this integration presents several issues because individual simulations are often designed for only a specific purpose and context. This leads to variations among space granularities and proposes a challenge when selecting an appropriate time granularity for the overall SoS simulation. To explore how this granularity might affect the outcome of simulations, we designed and developed a prototype system of three infrastructure simulation networks that were then combined into one SoS simulation using High Level Architecture (HLA) implementation. We then performed a series of experiments to investigate the response of the simulation to varying time granularities. Our examination included a propagation of disruptions among constituent simulations to estimate how this was affected by the frequency of updates between those simulations, i.e. time granularity. Our results revealed that the size of the simulated disruption decreased with in-creasing time granularity and that the simulated recovery time was also affected. In conducting this project, we also identified several ideas for future research that focus on a wider range of disruption generators and infrastructure systems in those SoS simulations.