Advances in Cross-Source Intelligent Security Monitoring and Resilience Strategies for Urban Cyber-Physical Power Systems: A Comprehensive Review

Abstract

The rapid integration of cyber-physical systems (CPS) into power grids has transformed traditional electricity networks into cyber-physical power systems (CPPS), enhancing efficiency but introducing profound cybersecurity vulnerabilities. Coordinated cyber-physical attacks, cascading failures, and information asymmetry exacerbate risks, as evidenced by incidents like the 2019 Venezuela blackout and disruptions during the Russia-Ukraine conflict. This review synthesizes advancements in cross-source intelligent security monitoring and analysis for power systems, focusing on urban CPPS models. Key technologies examined include multi-source log data aggregation for topology modeling, anomaly detection via self-supervised contrastive learning for missing node identification, tri-level optimization defenses against coordinated attacks, and vulnerability assessments using spectral metrics. We further discuss fault recovery strategies incorporating information delay models and graph convolutional networks (GCN) for optimal repair paths. Innovations such as multi-functional module integration (monitoring, powering, control) and asymmetric topology-aware attacks address gaps in high-security urban cores. By evaluating resilience metrics like load loss rates and islanding efficacy [35], this paper highlights pathways to enhance grid robustness, reduce outage costs, and support smart grid evolution. Future directions emphasize scalable AI-driven platforms for real-time threat mitigation, offering a comprehensive framework for researchers and engineers to bolster CPPS resilience.