CyberDep: Enhanced Generation of Bayesian Networks through the Inclusion of Bidirectional Data Flow Dependencies in Cyber-Physical Power Systems

Abstract

Power systems have been analyzed and studied as purely physical systems for a long time. Such efforts were critical to the establishment of the power grid as it is today. However, with the increased interest in the integration of renewable energy, the grid is experiencing more vulnerabilities to its operation, stability, and resiliency from the cyber realm. As such, it is crucial to understand the cyber-physical power system interdependencies. In this paper, we advance a Bayesian Network generation algorithm, called CyberDep. CyberDep quantifies cyber-physical interdependencies through conditional probability calculations and aids in analyzing bidirectional data flow dependencies and n-to-1 nodal connections between elements. CyberDep is implemented on a dataset of the cyber-physical emulation of the WSCC 9-bus system, which includes running physical, cyber, and cyber-physical disturbances on the system. The results showcase an improved interdependency quantification and visualization of the n-to-1 probabilistic relationships between the physical and cyber system components.