一种面向缺资料流域的三源特征相似性迁移水文模拟方法
By constructing a three-source feature space and feature mapping, and combining cluster analysis and supervised learning, the problems of single similarity measurement and coarse transfer strategy in data-scarce watershed hydrological simulation are solved. This enables refined transfer of hydrological model parameters and dynamic updating of the knowledge base, thereby improving the accuracy and adaptability of the simulation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- DALIAN UNIV OF TECH
- Filing Date
- 2026-05-19
- Publication Date
- 2026-07-17
AI Technical Summary
Existing hydrological simulation methods for watersheds lacking data suffer from problems such as a single similarity metric, a crude migration strategy, and an inability to update the knowledge base, resulting in inaccurate simulations of hydrological models when runoff observation data is unavailable.
A three-source feature space of geography, rainfall, and runoff is constructed. Through cluster analysis and feature mapping, the adaptive migration of hydrological model parameters and dynamic updating of the knowledge base are realized. A multi-dimensional similarity characterization and hierarchical migration strategy are adopted, and the runoff response characteristics are estimated by combining a supervised learning model. The weights are dynamically adjusted to improve the simulation accuracy.
It enables refined migration and continuous optimization of hydrological model parameters in watersheds with scarce data, significantly improving the reliability and generalization ability of the simulation and overcoming the limitations of existing technologies.
Smart Images
Figure CN122221702B_ABST