Nuclear power network security situation awareness and health degree assessment method and device
By using deep learning technology in nuclear power networks for feature extraction, timing analysis, feature matching and correlation analysis, combined with multi-task learning and integrated learning models, the problems of complex attack pattern recognition and health assessment in nuclear power networks are solved, and more efficient security situation awareness and health assessment are achieved.
Patent Information
- Application Number
- CN202411769236.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-04
- Publication Date
- 2025-03-28
AI Technical Summary
Due to the diverse data sources and complex characteristics in nuclear power networks, traditional security monitoring methods are difficult to deal with complex attack patterns and increasing security threats, resulting in limited accuracy and real-time nature of situational awareness.
Feature extraction is performed using a shared encoder, combined with long and short-term memory networks, gated recurrent units or timing convolutional networks for timing feature extraction, and constructing a feature matching layer by comparing learning matching traffic data and log data, introducing a self-attention mechanism for correlation analysis and feature fusion, using a multi-task learning model for event classification, and health assessment is performed through an integrated learning model.
It improves the accuracy and real-time awareness of nuclear power network security situations, can more effectively identify complex attack patterns and potential threats, provide more accurate health assessments, and support the security operations of nuclear power networks.