一种面向电能计量流水线的多分支故障检测方法及系统
By employing a multi-branch feature extraction network and knowledge graph verification method, the problems of inaccurate fault detection and poor adaptability of power metering pipeline equipment were solved, achieving higher detection accuracy and adaptability to new faults.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- FEIDONG COUNTY POWER SUPPLY CO STATE GRID ANHUI ELECTRIC POWER CO
- Filing Date
- 2026-04-07
- Publication Date
- 2026-07-17
AI Technical Summary
Existing fault detection technologies for power metering automated equipment suffer from insufficient fusion of multimodal features, inaccurate detection, poor adaptability to new fault features, and failure to fully utilize prior knowledge, resulting in a high false alarm rate.
A multi-branch feature extraction network is used to process multi-source heterogeneous sensing data, generate multi-dimensional representation features, and obtain enhanced joint features through adaptive gating network fusion. Fault features are verified by combining knowledge graph and classifier, and stable or emergency loops are selected to update network parameters to identify faults.
It improves the accuracy and adaptability of fault detection, enhances the ability to identify new faults, reduces the false alarm rate, and improves the intelligent detection capability of production line equipment.
Smart Images
Figure CN121980247B_ABST