Author(s): Yifan Yang; Yushu Xie; Julia C. Mullarney
Linked Author(s): Yifan Yang
Keywords: Flow-field reconstruction; Convolutional autoencoder; Physics-informed model; Coarse grains
Abstract: Flow measurements in open channel flows are essential for ascertaining both the overarching flow strength and turbulent characteristics. However, capturing measurements in energetic, and highly spatially and temporally heterogeneous environments is often logistically challenging and costly. In nowadays, digital surrogate-like tools are offering new solutions to those difficulties in water management. This study introduces a convolutional autoencoder-based neural network model for reconstructing real-time velocity and turbulent kinetic energy (TKE) fields over coarsegrain beds using boundary profiles. The overall aim is to establish an efficient and extensible model architecture for handling various input/output data formats for river flows.
Year: 2024