An efficient intelligent real-time monitoring and early warning method and system
By combining a lightweight YOLO v8 model and a lightweight expert model, the real-time and accuracy issues of arc monitoring in substation protection dead zones are solved, achieving efficient and intelligent arc fault monitoring and early warning, adaptable to remote and complex environments.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HUNAN POLICE ACAD
- Filing Date
- 2026-04-21
- Publication Date
- 2026-07-21
AI Technical Summary
Existing arc fault monitoring and early warning systems suffer from problems such as low monitoring accuracy, high false alarm rate, poor real-time performance, high consumption of computing resources, and poor communication conditions in substation protection dead zones, making it impossible to achieve efficient and accurate monitoring and early warning of arc faults.
A lightweight YOLO v8 model is used for real-time detection of fault arcs. The lightweight expert model is trained by distillation of a large general model and low-rank fine-tuning, and the arcs are preprocessed in a structured manner and risk assessment is performed to achieve full-process control from short-term disposal to long-term governance.
It achieves millisecond-level target detection and precise positioning, reduces the demand for computing resources, adapts to network outages and weak network environments, provides efficient and accurate arc fault monitoring and early warning, and improves the handling efficiency and intelligence level of operation and maintenance personnel.
Smart Images
Figure CN122435530A_ABST