Ontology-Driven Risk Modeling and Simulation Framework for Earthquake-Affected Complex Systems

Abstract

Earthquake-induced disruptions pose a significant threat not only to structural safety but also to the operational continuity and economic resilience of critical infrastructure. While conventional methods predominantly focus on physical damage estimation, this study focuses on translating established seismic parameters into digital representations that support comparable risk assessment outcomes in computational systems. In this study, we propose an ontology-driven, graph-based earthquake risk assessment framework integrated with a discrete-event operational simulation. Seismic parameters including fault mechanism, focal depth, soil classification, magnitude, and shaking duration are modeled as hierarchical entities within a graph database, enabling flexible semantic representation and efficient relationship traversal. A two-stage logarithmic-weighted risk calculation algorithm grounded in physical energy scaling laws is introduced to compute a normalized risk score. This score is subsequently coupled with a discrete-event operational simulation to quantify operational downtime, production loss, and financial impact at the facility level. Experimental evaluation across twenty seismic scenarios demonstrates a strong and consistent relationship between the proposed risk score and all operational impact metrics. The framework is further validated across four facility archetypes, confirming the risk score as a transferable relative indicator of operational impact whose absolute magnitude remains facility-specific. The results confirm that the proposed framework effectively translates seismological parameters into actionable operational insights, providing a robust decision-support tool for infrastructure resilience planning and seismic risk mitigation.