Distributed Simulation of Interdependencies in Community Resilience
Lin, Szu-Yun
2020
Abstract
When a disruption such as a severe natural event occurs, the interdependencies between the infrastructure systems of society can lead to cascading events that can adversely affect community resilience. Resilience is the ability of a community to withstand, adapt and recover from a disruption, typically measured in terms of loss of life, injuries and economic cost. Studying the interactions between infrastructure systems is complicated by the fact that each system is rooted in a specific field and thus requires crossing disciplinary barriers. To overcome the identified research challenge, this dissertation employs distributed computational simulation to model and investigate the interdependencies that arise during severe disasters and the post-disaster recovery process. It focuses in particular on multi-scale interdependencies and their time-dependent effects on community resilience. A distributed simulation framework that links each discipline specific simulator using a publish-subscribe data transmission pattern is proposed. The framework’s capabilities are demonstrated through a case study of multistory buildings that suffer wind-induced progressive damage. Data transmission is achieved using the Lightweight Communications and Marshalling (LCM) libraries. Building upon the LCM platform, a group of discipline specific computational models with disparate temporal and spatial scales are linked together to investigate the time-dependent interdependencies that arise between water, gas, and electrical power systems during a series of seismic events and the corresponding recovery processes. The results show that ignoring interdependencies can adversely affect resilience assessments and adopting time-varying recovery strategies can lead to better resilience performance. An agent-based computational model simulating benefit fraud behavior in the wake of a disaster is used to demonstrate that distributed simulation frameworks can take into account broader socio-technical interactions in resilience research. The study not only considered the effect of micro-level disaster-caused demands but also meso-level social factors on criminal tendencies. The proposed model captures the key characteristics of post-disaster benefit fraud in detail, including the dynamic nature of the criminal process. The results of parametric sensitivity analyses can be used to achieve a meaningful balance between the loss of fraudulent payments and the speed of distributing aid for improving the overall resilience performance of communities. To provide a scalable, versatile, and user-friendly solution for natural hazards simulations, a new distributed computing tool called Simple Run-Time Infrastructure (SRTI) is employed. The high-level structure, data structure, and fundamental components of SRTI are comprehensively described. The applications of SRTI in natural hazard simulations are presented. The performance of an initial version of the SRTI is compared with the LCM. A cross-language simulation of time-dependent resilience analysis of an electric power system is conducted to show the scalability and flexibility of the improved version of the SRTI, which reduces a user’s effort for composing a complex distributed simulation and better handle time management. Lastly, the choice between different versions of SRTI and potential features to develop in the future are discussed.Subjects
Community Resilience Distributed Simulation Interdependencies in Disaster Benefit Fraud in the Wake of Disaster Time-Dependent Resilience Assessment
Types
Thesis
Metadata
Show full item recordCollections
Remediation of Harmful Language
The University of Michigan Library aims to describe library materials in a way that respects the people and communities who create, use, and are represented in our collections. Report harmful or offensive language in catalog records, finding aids, or elsewhere in our collections anonymously through our metadata feedback form. More information at Remediation of Harmful Language.
Accessibility
If you are unable to use this file in its current format, please select the Contact Us link and we can modify it to make it more accessible to you.