PROBABILISTIC GRAPHICAL MODEL FOR ASSESSING THE RISKS OF DAMAGE TO CRITICAL INFRASTRUCTURE FACILITIES FROM DESTRUCTIVE PROCESSES

Abstract

In this work, the process of modelling the destructive processes, risk of their emerging, and damage probability was described. The modelling displayed a method to assess the risk of damage to critical infrastructure facilities. The destructive processes were independent but may be simultaneous, which is crucial to take into account while modelling. This paper presents a novel Bayesian network approach for assessing multi-risk critical infrastructure, focusing on the risks posed by bombardment and cyber threats. The model classifies risks into two levels: low and high, and uses these classifications to predict potential infrastructure damage. The Bayesian network is constructed with four nodes representing bombardment (B), cyber-threats (C), and damage (D) to infrastructure (I). This study presents a comprehensive risk analysis model for crisis management, which covers pre-crisis, response, and post-crisis stages, evaluating risk distinctly in each. The model hypothesises that risk fluctuates based on multi-hazard processes and can be evaluated for each vulnerable object. Risk dimensions include the probability of threat, target attributes, and accessibility to threat-creating actors. A multi-hazard risk assessment is represented as a combination of these components, providing a dynamic risk rating for each entity. This aids in informed decision-making throughout the crisis management cycle and allows for a comprehensive understanding of the system’s risk profile, taking into account the complex interplay of different risk factors. It provides a robust framework for risk assessment in complex systems and offers valuable insights for decision-makers in critical infrastructure protection, enabling them to make informed decisions about resource allocation, risk mitigation strategies, and emergency response planning.