A Fault Diagnosis Method of Power Networks Including Information Attacks

Modern smart grid is an electrical cyber-physical system with physical power network and interconnected communication network as the core, and its rich information interconnection characteristics increase its risk of cyber attacks. This kind of fault information attack is very hidden, extremely destructive, and can perfectly evade existing fault diagnosis methods. Therefore, this paper proposes a power grid fault diagnosis method considering the attack of fault information. This method has two modules, a fault diagnosis module based on remote signal data and a fault diagnosis module based on telemetry data, and the two modules run in parallel to make comprehensive decisions. The remote signaling module uses the remote signaling data scrubbing correction model based on the Associative Memory Neural Network (AMNN) to eliminate redundant amounts and correct the erroneous remote signaling data that has been tampered with. The key remote signaling coding set is mapped to the global diagnostic space, and the fault components and fault classification conditions are output by the minimum distance classification method to complete the diagnosis of the remote signaling module. The telemetry module proposes a remote survey attack identification and repair method based on fuzzy C-means clustering of the number of two-layer optimal clusters, which completes the detection and correction of the attacked data. Then, Ensemble Empirical Mode Decomposition (EEMD) is used for fault feature extraction. The Neuron Study Forgetting Reasoning Algorithm (NSFRA) was used to establish a pulsed neuromembrane system for learning forgetting reasoning, and decision-making reasoning was performed according to the Neuron Matrix Reasoning Algorithm (NMRA) to complete the telemetry module diagnosis. Finally, the diagnostic results of the above two modules are entered into the comprehensive decision module for comprehensive diagnosis to complete the final decision. The method was verified using the IEEE39 bus system.