Probabilistic Framework for Evaluating Community Resilience: Integration of Risk Models and Agent-Based Simulation

Abstract

This paper proposes a novel probabilistic framework to quantitatively evaluate the resilience of communities comprising buildings and various interdependent infrastructure systems. To this aim, the proposed framework seamlessly integrates risk models and agent-based simulation in a Monte Carlo sampling scheme. The risk module includes models that evaluate the initial posthazard state of the community by probabilistic simulation of the hazard event, the structural response and damage of buildings and infrastructure systems, and cascading consequences that arise from interdependencies. Subsequently, the agent-based module simulates the recovery of the community from those consequences in which decentralized autonomous decision-making entities called “agents” undertake recovery operations. The agents prioritize buildings and infrastructure components for recovery and schedule operations as discrete events with uncertain duration and cost. Consequently, the probability distribution of the total cost incurred by the community and the total recovery time is evaluated. A resilience measure is then proposed as a function of the total community cost, which represents demand, and the gross regional product of the community, which represents the capacity to cope with that demand. The framework is showcased by a comprehensive application to a community comprising a portfolio of residential and commercial buildings, an electric power system, a water system, and a healthcare system subject to seismic hazard.