CBR method for risk assessment on power grid protection under natural disasters: Case representation and retrieval

Abstract

Risk analysis is always the pivotal part of emergency preparedness for critical infrastructure protection such as power grid and traffic network. The main contribution of this paper is to employ case-based reasoning (CBR) method (combines case representation and retrieval) to illustrate a risk assessment framework for protecting power grid. It focuses on two core key parts: (1) Using ontology model to express precursors of risk, the described semantic network contains sub-concepts to outline selected precursors from hazards, environment, responders and physical system and (2) by analysis of emergency scenario precursors with sub-concept similarity and fuzzy value similarity calculation, the potential risks could be recognized to assist to retrieve the past knowledge, and effective and feasible actions would be taken to decrease the threats or cut disaster loss for power grid. Via a case study, the result shows that the proposed method extends the applicability of conventional CBR technique to numbers of real-world settings.