Abstract
Improving the accuracy of the vulnerability assessment of the power cyber-physical system has important practical value for protecting the vulnerable links of the system and preventing large power outages. This article studies the vulnerability assessment of power cyber-physical systems under cyber-attacks. First, from the perspective of topology, the electrical betweenness is used as the structural vulnerability index of the power system. Based on the power flow characteristics of the power network, the voltage is used as the state vulnerability index. Then, starting from the structure, the node degree and clustering coefficient are selected to analyze the structural vulnerability. Considering the service transmission characteristics of the information system, the power service importance is selected as the vulnerability assessment index of the information system. The two selected indexes are used to construct a quantitative formula, and a comprehensive index is obtained to complete the node vulnerability assessment of the power information system. Then, on the basis of the system vulnerability assessment, we screen to determine the entry and target nodes of network attacks and predict the network attack path. Finally, the effectiveness of the proposed method for vulnerability assessment and feasibility is verified by MATLAB programming simulation.
Introduction
Modern power grids have evolved into complex Cyber-Physical Systems (CPS), where physical power delivery is tightly integrated with advanced communication and control networks. While this digital transformation enhances efficiency and reliability, it also introduces significant cyber vulnerabilities. Traditional assessment methods often analyze the power grid and information network in isolation, failing to capture the cascading effects of cross-domain failures. This research addresses this gap by proposing a unified vulnerability assessment framework that considers both structural topology and operational physics.
Methodology: A Multi-Layered Approach
Physical Layer Vulnerability
The physical layer assessment focuses on two dimensions: structural and state vulnerability.
- Structural Vulnerability (Electrical Betweenness): Standard graph theory metrics like "shortest path" betweenness are inadequate for power grids because electricity follows Kirchhoff’s laws rather than simple topological paths. The authors utilize electrical betweenness to measure the importance of a branch based on unit injection currents between generator-load pairs.
- State Vulnerability (Voltage Stability): Using a Modified Nonlinear Power Flow (PNFA) algorithm, the system evaluates how close each node is to its voltage stability limit, identifying nodes at risk of collapse under stress.
Cyber Layer Vulnerability
The information system is assessed based on its network structure (node degree and clustering coefficients) and the criticality of the services it carries.
- Service Importance: Data packets are weighted by their function. For instance, "Relay Protection" signals are prioritized over "Office Automation" data, reflecting their critical role in grid stability.
Attack Path Prediction
By identifying the most vulnerable cyber entry points and critical physical targets, the framework can predict potential attack trajectories. The authors employ Dijkstra’s algorithm to map the most likely paths an attacker would navigate through the coupled network, where "distance" is defined by ease of exploitation and potential system impact.
Case Study: IEEE 14-Bus System
The proposed model was validated using the standard IEEE 14-bus system. By mapping a corresponding 14-node information network, the researchers demonstrated that:
- Physics-Aware Metrics Matter: Electrical betweenness identified high-risk lines (e.g., lines 4–9 and 5–6) that traditional capacity-based models overlooked.
- Critical Node Identification: Node 1 (the dispatch center) emerged as the most critical due to its central role in information flow and business importance.
- Transit Nodes as Choke Points: The simulation highlighted that certain transit nodes in the information layer serve as critical bottlenecks, requiring enhanced security hardening.
Conclusion
This research provides a comprehensive framework for securing modern smart grids. By fusing physical power flow characteristics with cyber network topology, the model offers a more accurate assessment of system vulnerability than traditional isolated methods. Future work will focus on scaling these algorithms for larger real-world networks and incorporating dynamic load changes into the vulnerability analysis.
Summary updated