Substation defect identification method and system based on unmanned aerial vehicle inspection data
By using multi-sensor data fusion and multi-model decision-making techniques, a substation defect identification model was constructed, which solved the problem of poor detection performance of single sensors in UAV inspections and achieved high reliability and high accuracy defect identification under different lighting conditions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- STATE GRID HUNAN ELECTRIC POWER COMPANY LIMITED
- Filing Date
- 2024-06-24
- Publication Date
- 2026-07-21
AI Technical Summary
Existing methods for identifying defects in substations using drones rely on a single sensor, resulting in poor detection performance under poor lighting conditions or when the substation is well-hidden. Furthermore, the analysis depends on expert experience, leading to poor reliability and accuracy.
A multi-sensor data fusion method is adopted, which combines acoustic print, vibration, visible light and infrared image data. A substation defect identification model is constructed by using Transformer network, LSTM network and YOLOv4 model. The final identification is performed by multi-model decision fusion technology.
This improves the reliability and accuracy of substation defect identification by drone inspections, enabling accurate identification of substation defects under different lighting conditions.
Smart Images

Figure CN118710967B_ABST