Loading...

Optimal Enhancement of Community Resilience through Prioritization of Recovery Operations for Interdependent Infrastructures

Naderi, Mahdi | 2019

651 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 52029 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Mahsuli, Mojtaba
  7. Abstract:
  8. In this study, community resilience after a disaster is optimally enhanced for interdependent infrastructures through prioritization of recovery operations considering limited resources. Optimal enhancement of community resilience can be performed at two levels: Pre-disaster and Post-disaster. In the pre-disaster level, resilience enhancement is achieved through optimization of long-term decision making that includes optimized allocation of resources to retrofit of buildings. However, in the post-disaster response and recovery period, allocating resources to retrofit or backup systems is infeasible. Therefore, in the post-disaster level considered in this study, resilience enhancement is achieved through optimal prioritization of resource allocation for recovery operations of infrastructures and optimal prioritization of search-and-rescue operations. Indeed, the recovery process is enhanced through prioritizing the recovery operations of those infrastructure components whose lack of functionality leads to catastrophic cascading losses. The objective function of the aforementioned optimization problem is a resilience index calculated using the Rtx resilience evaluation framework, which has been developed at Sharif University of Technology. This index is computed using the total cost incurred by the community under disaster. The probability distribution of the total community cost is computed using Monte Carlo sampling. It includes the costs associated with the repair and replacement of damaged structures and infrastructures, injuries and deaths, disruption of service, business interruption, life quality reduction due to injuries, temporary shelter, and other consequences. These costs are modeled through agent-based simulation during the recovery period of community. The present study maximizes the resilience index by optimizing the priorities of recovery operations of infrastructures. These priorities are determined as integer numbers. Indeed, maximizing the resilience index leads to minimizing the total cost incurred by the community. The genetic algorithm method is employed to conduct optimization. The optimal enhancement of resilience has been the subject of several previous studies. However, most studies either employed linear programming methods through simplification to solve such a highly nonlinear problem or ignored the uncertainties in this highly uncertain issue. Other similar studies have either concentrated only on optimization of one infrastructure or have ignored the dependency of infrastructure components. The present study is novel in that it approaches the problem through decentralized optimization based on the probabilistic simulation of events that occur during the recovery period of interdependent infrastructures. For example, unserviceability of a power station leads to a capacity reduction of dependent hospitals in the treatment of injuries, not to mention the business interruption of dependent commercial organizations. Furthermore, this study comprehensively accounts for the cascading consequences. In this framework, in addition to direct economic and social losses, costs incurred by the community due to an elongated recovery process of infrastructures is considered in the resilience index and consequently, in the optimal enhancement of community resilience. Another novelty of this study is the utilization of decentralized optimization for more accurate modeling of the decision-making process. These decisions are made by independent agents that autonomously recover their designated infrastructures, and there is no central authority that issues directions to the agents to optimize the recovery of the entire community. The methodology presented in this study is showcased by a case study for a hypothetical region in Tehran. The optimization results indicate a 2% reduction in the total cost incurred by the community leading to $3 million in savings. This enhancement is achieved without using new resources and only by optimizing the recovery process. This framework, as a decision support system, is capable of evaluating the effect of different decisions on the community resilience, and optimize these effects by introducing optimal decisions
  9. Keywords:
  10. Community Resilience ; Distributed Optimization ; Linear Integer Optimization ; Recovery ; Interdependent Infrastructure Systems ; Consequence Analysis ; Monte Carlo Sampling ; Cost Analysis ; Disaster Management

 Digital Object List

 Bookmark

No TOC