Power system source-load disturbance scene generation method in extreme weather scene
By constructing a generative adversarial network model and combining spatiotemporal correlation learning and multi-head attention mechanism, source-load disturbance scenarios under extreme weather conditions with high realism and diversity are generated, which solves the problem of insufficient scenario generation in existing technologies and improves the risk assessment and disaster prevention and mitigation capabilities of power systems.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- STATE GRID SHANGHAI MUNICIPAL ELECTRIC POWER CO
- Filing Date
- 2025-12-19
- Publication Date
- 2026-05-15
AI Technical Summary
Existing technologies lack methods to accurately capture spatiotemporal correlations and effectively generate highly realistic source-load disturbance scenarios under extreme weather conditions, resulting in insufficient risk assessment of power systems under extreme weather conditions.
By constructing a generative adversarial network model and combining spatiotemporal correlation learning and multi-head attention mechanisms, the generator can automatically extract and learn the complex spatiotemporal variation patterns and dependencies of extreme weather events, generating highly realistic and diverse source-load disturbance scenarios.
It improves the accuracy and realism of scene generation, enhances the ability to generate extremely scarce samples, provides a more comprehensive and reliable risk assessment basis, and improves the disaster prevention and mitigation capabilities of the power system under extreme weather conditions.
Smart Images

Figure CN122045803A_ABST