Attack-resilient state estimation for cyber-physical power systems: A dynamic spatial-temporal redundancy reconfiguration framework for FDIA detection

Modern power systems, as cyber-physical systems, increasingly rely on hybrid measurement data to improve the accuracy and resolution of state estimation (SE). However, the enhancement of SE functionality is accompanied by an increased reliance on measurement devices and external detection mechanisms, thereby expanding the attack surface and exposing SE to sophisticated cyber threats. This paper reveals security vulnerabilities in existing hybrid measurement-based SE frameworks, particularly under coordinated false data injection attacks (FDIAs) that manipulate both baseline and verification measurements to evade detection. To address this challenge, we propose an attack-resilient SE method based on dynamic spatial-temporal redundancy reconfiguration. By proactively injecting measurement uncertainty into the measurement process, the method enhances resilience against external attacks. A comprehensive detection index is introduced to jointly evaluate estimation accuracy and attack impact. Then, we develop an FDIA detection framework that integrates offline training and online adaptation. The offline phase optimizes the sensitivity parameter and initial measurement configurations, while the online phase dynamically updates measurement reconfiguration strategies and detection thresholds based on real-time feedback. Extensive validations on the IEEE 14-bus and 118-bus systems demonstrate that the proposed approach significantly improves the FDIA detection capability while maintaining estimation stability and computational efficiency, without requiring additional external security mechanisms.