Security Resilience Assessment of Industrial Control Network System for Pipeline Network Based on CN-IRML

In response to the cybersecurity threats and insufficient system resilience faced by industrial control networks in pipeline systems, this paper proposes a security resilience assessment framework that integrates complex network theory with static integrated resilience machine learning (CN-IRML). By constructing a topological model of the pipeline industrial control system, this framework quantifies the structural characteristics of nodes and edges, and analyzes network vulnerabilities based on typical attack scenarios. Using the IRML model, it extracts static features and calculates resilience metrics for critical nodes under constraints of interference time, buffer time, and recovery time, thereby classifying risk levels. Case studies demonstrate: Taking a certain oil and gas pipeline compression station with a design capacity of $6 \times 10^{8} \mathbf{S m}^{3} /$ a as an example, high-risk nodes such as the G3 separator and G4 pressure regulating device are identified, with betweenness centrality $\mathbf{\gt 0.5}$; their failure would result in a $\mathbf{4 0 \%}$ or more reduction in transmission capacity. After redundancy upgrades, system resilience significantly improves, with recovery time reduced to within $\mathbf{1 0}$ hours.