Application of PINNs to Define Roughness Coefficients for Channel Flow Problems

Применение PINN для определения коэффициентов шероховатости в задачах течения в каналах
Sergei Vladimirovich Strijhak, Konstantin Borisovich Koshelev, Андрей Геннадьевич Болотов
2025-09-16

Delft3DInverse roughness coefficient estimationOpen-channel flowPhysics-informed neural networksSaint-Venant equations
This paper considers the possibility of using Physics-Informed Neural Networks (PINNs) to study the hydrological processes of model river sections. A fully connected neural network is used for the approximation of the Saint-Venant equations in both 1D and 2D formulations. This study addresses the problem of determining the velocities, water level, discharge, and area of water sections in 1D cases, as well as the inverse problem of calculating the roughness coefficient. To evaluate the applicability of PINNs for modeling flows in channels, it seems reasonable to start with cases where exact reference solutions are available. For the 1D case, we examined a rectangular channel with a given length, width, and constant roughness coefficient. An analytical solution is obtained to calculate the discharge and area of the water section. Two-dimensional model examples were also examined. The synthetic data were generated in Delft3D code, which included velocity field and water level, for the purpose of PINN training. The calculation in Delft3D code took about 2 min. The influence of PINN hyperparameters on the prediction quality was studied. Finally, the absolute error value was assessed. The prediction error of the roughness coefficient n value in the 2D case for the inverse problem did not exceed 10%. A typical training process took from 2.5 to 3.5 h and the prediction process took 5–10 s using developed PINN models on a server with Nvidia A100 40GB GPU.
1
Fully connected PINNs were applied to approximate 1D and 2D Saint-Venant equations for modeling river-channel hydrodynamics.
2
PINNs were used to solve the inverse problem of estimating channel roughness coefficients from flow-related data.
3
The 2D inverse estimation error for the roughness coefficient n did not exceed 10%; training required 2.5–3.5 hours, and prediction required 5–10 seconds on an Nvidia A100 GPU.
4
The models predicted velocity, water level, discharge, and wetted cross-sectional area in one-dimensional channel-flow problems.
5
Two-dimensional PINNs were trained on synthetic velocity and water-level data generated with Delft3D, while hyperparameter effects on prediction quality were evaluated.

Model river-channel flow sections governed by the Saint-Venant equations, including 1D rectangular channels and 2D channel-flow cases

Hydraulic-state prediction and inverse estimation of channel roughness coefficient n, including velocity, water level, discharge, and cross-sectional area, using PINNs

Publication Details
Publication Date
2025-09-16
Journal
Publisher
ISSN
Cited by
3
Access Type
Author Information
Authors
Sergei Vladimirovich Strijhak
Konstantin Borisovich Koshelev
Андрей Геннадьевич Болотов
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%