一种光伏系统状态融合决策方法及系统
By introducing physical mechanisms in real time through cascaded sensing networks, the problem of the separation between data-driven and physical mechanisms in photovoltaic system condition monitoring is solved, enabling highly reliable and interpretable photovoltaic system condition monitoring decisions and generating decision-making basis with physical semantics.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 国网安徽省电力有限公司营销服务中心
- Filing Date
- 2026-03-26
- Publication Date
- 2026-07-17
AI Technical Summary
Existing photovoltaic system condition monitoring methods suffer from problems such as the disconnect between data-driven models and physical mechanisms, insufficient model interpretability, and superficial integration schemes, resulting in a lack of high reliability and traceability in the decision-making process.
A state fusion decision-making method for photovoltaic systems based on real-time guidance and self-explanation using physical mechanisms is adopted. Multimodal data processing is performed through a cascaded sensing network, physical mechanisms are introduced in real time and hierarchical interpretable evidence is generated, and deep fusion feature vectors and decision results are output.
It achieves high reliability and interpretability of photovoltaic system condition monitoring model, improves the robustness of decision-making process and operation and maintenance practicality, and can correct the reasoning bias of data-driven deviating from physical common sense under complex and abnormal working conditions, and generate decision basis with physical semantics.
Smart Images
Figure CN121920558B_ABST