Artificial Intelligence Based Network Security Threat Detection and Response Mechanism for Electric Power System

Artificial Intelligence Based Network Security Threat Detection and Response Mechanism for Electric Power System

연구 분야: Safety

학회: 2025 International Conference on Electrical Drives, Power Electronics & Engineering (EDPEE)

The expansion of electric power system (EPS) network leads to the complexity of attacks, and traditional protection can’t meet the security requirements. The purpose of this paper is to study the security threat detection and response mechanism of EPS network based on artificial intelligence (AI), so as to improve the security protection capability of EPS. In order to achieve this goal, a threat detection system based on neural network is first constructed. The system can automatically extract the characteristics of multi-source data such as network traffic and system logs, and identify potential network threats in real time. A comprehensive empirical study is carried out in the actual EPS network environment, and the detection performance of the model is verified by designing various attack scenarios. The results show that the model has achieved excellent performance in key indicators such as accuracy, recall and F1 value. The detection speed of the model is fast, the false alarm rate and the false alarm rate are low, and the consumption of system resources is also acceptable. This proves the practicability of threat detection and response mechanism based on AI in the field of EPS network security.