A ship lock chamber water level fluctuation analysis control method based on deep learning
By using deep learning to predict water level and flow velocity, and combining shallow water equations and momentum equations, the valve action rate and scheduling step size are dynamically adjusted, solving the problem of predicting water level fluctuations in the lock control system and improving the safety and efficiency of lock operation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- TIANJIN RES INST FOR WATER TRANSPORT ENG M O T
- Filing Date
- 2026-05-20
- Publication Date
- 2026-07-17
AI Technical Summary
Existing lock control systems are unable to predict and suppress local water level fluctuations in advance. Fixed scheduling steps lead to high computational loads and ineffective mechanical wear. Pure data-driven models lack physical boundary constraints, posing safety risks.
A deep learning-based approach is adopted to predict water level and flow velocity through a physical information long short-term memory network, calculate the momentum equation residual by combining shallow water equations, dynamically calculate the upper and lower limits of valve action rate, and use a policy neural network to generate control commands within the dynamic safety boundary and dynamically adjust the scheduling step size.
When a turbulent flow trend is predicted, the system automatically suppresses large fluctuations in water level, improves the safety and water surface stability of the lock filling and emptying process, reduces the computational load and mechanical wear of the controller, and enhances the engineering reliability of the control system.
Smart Images
Figure CN122411601A_ABST