Cascading Failure Risk Assessment Based on Event-Driven Model in a Cyber-physical Power System

The integration of communication technology into power systems enhances their observability and controllability but also introduces additional risks caused by the interaction between cyber and physical systems, leading to the expansion of cascading failures in cyber-physical power systems (CPPSs). To address this problem, this article proposes a cascading failure risk assessment method for CPPS based on an event-driven model, which provides technical support for risk warning and security analysis of CPPSs. The proposed method involves the establishment of a state transition matrix of CPPS based on a dynamic Bayesian network and analysis of the triggering law of cascading faults. A dynamic HMM model and a set of expected risk scenarios of CPPS are established based on the observable results of physical system faults by the information system. Furthermore, the corresponding multi-index risk assessment model of CPPS is established to describe the risk propagation ability of the system in its operation state, followed by the migration of the operation mode according to the adjustment of the high-risk event-driven control strategy. Finally, the effectiveness of the method is verified using the IEEE-39 system.