Resilience Evaluation of Cyber-Physical Power Systems Based on Evolutionary Algorithms

Limiao Zhang, Xilong Zhou, Haiping Ma, Ye Tian, Hai-Feng Zhang, Xudong Huang, Xingyi Zhang

Abstract: Cyber-physical power systems (CPPSs) significantly improve the power grid’s performance but also expose new security threats. Malicious attacks on individual components can potentially lead to catastrophic power outages. With the emergence of new hacker methods, it is necessary to foresee the destructive attacked targets and design effective recovery strategies in advance. In this article, we propose a two-stage framework for evaluating the resilience of CPPSs based on evolutionary algorithms (EAs), which consists of an “attack” phase and a “recovery” phase. To simulate real-world attacks, we determine the sequential attacked targets by solving a multiobjective optimization problem that balances the attack effect and costs. An ac-based cascading model is used to evaluate the CPPS performance degradation after being attacked. Then, we design corresponding recovery strategies to enhance system performance, considering both final and process recovery metrics. The framework is tested on a synthetic CPPS based on the IEEE 118 Bus System and the IEEE 300 bus system, separately. Our results show that, compared to traditional strategies based on topological centrality, CPPS is more vulnerable to EA-based attacks, and the EA-based strategy also outperforms in terms of system recovery, demonstrating the potential applicability of the framework in improving the resilience of real-world power systems against sophisticated cyber-physical threats. The proposed framework provides a basis for integrating advanced optimization techniques into security-aware system design, which can be applied as a decision-support tool for offline resilience planning and emergency response analysis in utility control centers.

External IDs:doi:10.1109/tr.2025.3639096