CO-EVOLUTION OF DISASTER RESILIENCE IN COMPLEX SOCIAL SYSTEMS

Post disasters, when designing disaster response and business continuity plans, stakeholders should consider various interdependencies among different organizations and sociotechnical systems for the society's effective and prompt restoration. This study proposes a co-evolutionary simulation for optimizing disaster responses adopted by different stakeholders using an agent-based model and a social system modeling framework that describes multiple interdependencies among lifeline infrastructures, industry activities, and civil life. In the simulation, the whole society co-evolved to swiftly recover after a disaster while interacting with various stakeholders' disaster responses, focusing on two stakeholder types: lifeline operators and industrial companies. Both lifeline restoration and backup supplier plans were simultaneously optimized via simulation using genetic algorithm (GA). Results show that GA worked appropriately; observations of the recovery processes of both types of stakeholders revealed that the proposed co-evolutionary simulation can coordinate various responses and help one understand the society's resilience management.