Vulnerability analysis of cyber-physical power systems based on failure propagation probability

This paper conducts vulnerability analysis on various nodes of the cyber-physical power systems based on failure propagation probability. An interdependent network model is established by considering the network topology of the communication network and the electrical characteristics of the power network. The model sets two failure propagation probabilities to describe the ability of failure propagation within and between the power and communication networks. These two probabilities quantify the likelihood of survival when a failure occurs in the branches and overloaded branches connected to the faulty nodes in the network, rendering the coupled network weakly dependent. A method is proposed to use the clustering coefficient of a node in a communication network as its load capacity. A linear programming method is chosen to schedule the output of power network generators and the load capacity of load nodes. The impact of four different network topologies on the overall system vulnerability of the models is also analyzed. Furthermore, the effectiveness of the linear programming economic dispatch method is verified by using the IEEE 39-node system as the power network of the weakly coupled network.