Power grid icing dynamic early warning method based on multi-modal feature fusion and physical guidance
By employing a multimodal feature fusion and physical guidance approach, the problems of insufficient data quality and inadequate model fusion in transmission line icing early warning were solved, achieving high-precision and dynamic icing early warning and risk assessment, and supporting power grid anti-icing dispatch.
Patent Information
- Application Number
- CN202610488752.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-04-14
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies struggle to acquire high-quality micro-scale meteorological characteristics of transmission line corridors, lack effective fusion of multi-source heterogeneous information and dynamic prediction models, resulting in low accuracy of power grid icing warnings and a lack of dynamic warning mechanisms guided by physical laws.
A multimodal feature fusion and physical guidance approach is adopted. Data repair and feature fusion are performed using Regional Meteorological Adaptive Graph Neural Network (RCAGNN) and Multi-Physical Condition Attention Constrained Generative Adversarial Network (MPAC-GAN). Combined with power grid operation status and satellite remote sensing microphysical products, a dynamic risk assessment system is constructed to generate high-resolution icing early warning.
It enables precise and dynamic early warning of icing on transmission lines, improves prediction accuracy and physical reliability, provides detailed risk location, cause analysis and process prediction, and supports power grid anti-icing dispatch.
Smart Images
Figure CN122413285A_ABST