A multi-modal anomaly confidence evaluation and hierarchical disposal method in a thermal power plant inspection scene
By using multimodal data fusion technology, the problems of false alarms and missed alarms in the anomaly assessment during thermal power plant inspections have been solved, enabling accurate risk assessment and differentiated handling of equipment anomalies, and ensuring the safe and stable operation of thermal power plant equipment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HEBEI HUADIAN SHIJIAZHUANG THERMOELECTRICITY
- Filing Date
- 2026-03-27
- Publication Date
- 2026-06-23
AI Technical Summary
In existing thermal power plant inspection technologies, there are problems of false alarms and omissions in anomaly assessment and handling, and the risk level classification is not accurate enough, making it difficult to achieve differentiated and graded handling.
Multimodal data fusion technology is adopted to collect multi-source heterogeneous modal data, extract defect geometric features and deterioration trend features, combine them with the fault correlation mechanism of thermal power equipment to perform feature fusion and weight optimization, and construct a multi-dimensional judgment matrix for abnormal risk assessment and graded handling.
It has improved the anti-interference capability and accuracy of anomaly assessment, realized the refined classification and handling of equipment anomaly risks, and ensured the safe and stable operation of key equipment in thermal power plants.
Smart Images

Figure CN122264436A_ABST