低压配电网表码值的异常检测与数据修复方法、装置、设备、存储介质和程序产品
By employing a three-tier cascaded architecture and intelligent attribution analysis, the problems of false alarm rate and missed alarm rate in the detection and repair of abnormal meter code values in low-voltage distribution networks were solved, achieving efficient data repair and quality improvement.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA SOUTHERN POWER GRID DIGITAL GRID GRP CO LTD
- Filing Date
- 2026-05-22
- Publication Date
- 2026-07-17
AI Technical Summary
In existing technologies, the methods for detecting abnormal meter readings in low-voltage distribution networks have high false alarm and false negative rates, and the data repair lacks the ability to attribute abnormalities, resulting in poor data repair performance.
An anomaly detection is performed using a three-layer cascaded architecture consisting of a statistical rule layer, a machine learning layer, and a physical constraint layer. Intelligent attribution analysis is then performed by combining multi-dimensional feature vectors and event log data to select a scenario-based repair method.
It reduced the false alarm rate and false negative rate of anomaly detection, achieved efficient data repair, and improved the quality and accuracy of table code values.
Smart Images
Figure CN122241545B_ABST