一种基于时空特征融合与先验知识约束的浓雾天气形势识别方法
By constructing a dense fog weather pattern identification method based on multi-time and multi-layer weather field tensors and prior attribute labels, the problem that dense fog identification in existing technologies is difficult to reflect the continuous evolution of weather processes and lacks meteorological semantic correspondence is solved, and more stable and interpretable dense fog forecast assistance is achieved.
CN122196506BActive Publication Date: 2026-07-17NANJING UNIV OF INFORMATION SCI & TECH
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NANJING UNIV OF INFORMATION SCI & TECH
- Filing Date
- 2026-05-12
- Publication Date
- 2026-07-17
Smart Images
Figure CN122196506B_ABST
Abstract
本发明公开了一种基于时空特征融合与先验知识约束的浓雾天气形势识别方法，属于气象信息处理与智能预报辅助技术领域。包括：构建过程级样本；根据过程级样本进行编码，并结合时间演变关系生成对应于所述天气过程的时空表示；对过程级样本进行编码，得到知识增强表示，并输出所述天气过程在各先验气象属性上的属性响应；对时空表示与知识增强表示进行特征融合，输出天气形势分类预测结果；建立一致性约束，以对识别模型的训练过程进行引导来训练识别模型。本发明能够将天气过程演变信息、区域背景场信息与地方判识经验进行统一建模，提高复杂浓雾背景自动识别的准确性、稳定性和可解释性，并为浓雾预报辅助提供结构化的背景提示信息。
Need to check novelty before this filing date? Find Prior Art