Modeling and Risk Assessment of Cyber Attacks in Distribution Grid Cyber-physical Systems

This paper proposes a risk assessment model for information-oriented attacks on power distribution network cyber-physical systems(CPS), which measures the comprehensive security risks of power distribution network CPS under Bayesian graph. The model consists of two key components: a network attack success probability model based on Bayesian graph, and an evaluation index for the physical consequences caused by cyber attacks. Firstly, several common attack methods in information systems are analyzed, and Bayesian attack graphs are used to model and quantitatively evaluate the probability of success of network attacks. Then, using Python to call PSS/E to establish a distribution network CPS system to simulate cascading failures caused by cyber attacks on power systems. Using the network fluctuation-frequency and voltage deviation caused by cyber attacks to establish physical consequence evaluation indicators, combined with the probability of cyber attacks, a comprehensive risk quantification evaluation model for the system is constructed. Finally, based on the simulation of an IEEE-39-node distribution CPS system, the feasibility and effectiveness of the proposed model are verified, and it is confirmed that the vulnerability of the system is positively correlated with the attack probability and physical consequences.