一种模型耦合的湖区水位预测方法、介质及计算机设备
By employing a model coupling method in lake level prediction, and utilizing the dynamic coupling of neural networks and two-dimensional hydrodynamic models, the problems of downstream boundary condition updates and rigid representative point configurations are solved, achieving high-precision and flexible lake level prediction that can adapt to extreme events and changes in boundary conditions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHANGSHA UNIVERSITY OF SCIENCE AND TECHNOLOGY
- Filing Date
- 2026-04-30
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for predicting water levels in lake areas suffer from problems such as the inability to dynamically update downstream boundary conditions, lack of a measured data feedback mechanism, redundant output dimensions, and rigid configuration of representative points, leading to the accumulation of simulation errors, insufficient prediction accuracy, and insufficient flexibility.
A model coupling method is adopted, which establishes a water level prediction neural network model in the PyTorch framework and couples it with a two-dimensional mathematical hydrodynamic model. The downstream boundary conditions are dynamically updated using real-time upstream water level data. Numerical simulation is carried out in combination with shallow water equations, and a lightweight neural network model is established for key points to predict water levels.
It achieves physical consistency and real-time correction of lake water level prediction, reduces computational redundancy and overfitting risk, improves the prediction accuracy and flexibility of key points, and adapts to extreme events and sudden changes in boundary conditions.
Smart Images
Figure CN122113696B_ABST