Reliability Assessment of Cyber-Physical Generation System

This paper presents an effective reliability assessment technique for the cyber-physical generation systems (CPGS)s. The assessment technique incorporates cybersecurity issues in smart grids considering non-normal random variables with nonlinear dependencies. A Bayesian attack graph is adopted for vulnerability analysis of CPGS and model the probabilistic nature of attack paths. In addition, the meantime-to-compromise is modeled at the operation level instead of the component level, which leads to more accurate results in terms of loss of load expectation and expected energy not supplied. Further, a hybrid approximate-analytical method is presented to assess the capacity outage probability table in power systems. The proposed method overcomes the Sequential Monte Carlo simulation (SMCS) drawbacks and demonstrates the effects of cyber-attack and wind farm correlations on generation capacity adequacy assessment. The proposed method is implemented on the modified IEEE Reliability Test System. A comparison with SMCS demonstrates the efficiency and accuracy of the proposed method.