Power operation and maintenance personnel identity authentication method based on biometric recognition
By using a biometric identification technology that combines infrared imaging and bone conduction vibration sensors in the live-line maintenance scenario of ultra-high voltage substations, and adjusting the modal fusion ratio in real time, the problem of authentication failure caused by anti-arc mask and electromagnetic pulse interference was solved, and highly reliable identity authentication was achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- LUOHE POWER SUPPLY OF HENAN ELECTRIC POWER CORP
- Filing Date
- 2026-04-17
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional biometric identification technologies fail in live-line maintenance scenarios at ultra-high voltage substations due to arc-proof masks and strong electromagnetic pulse interference. Existing multimodal fusion solutions fail to effectively utilize cross-modal physiological collaborative constraints, resulting in insufficient authentication robustness.
The study uses infrared imaging that can penetrate the mask coating and bone conduction vibration sensors to simultaneously collect periorbital vein texture and skull voiceprints. A dynamic weight matrix is generated through a pre-trained environment assessment network to adjust the modality fusion ratio in real time, and a twin network is used for identity matching.
Achieving seamless and robust continuous identity verification in extreme environments avoids authentication failures caused by damage to a single modality, thus improving the reliability of identity authentication for power operation and maintenance personnel.
Smart Images
Figure CN122416539A_ABST