Hydropower prediction using a new Physics-informed neural network
In conjunction with the Marine Renewable Energy Technology group at Natural Resources Canada, and the Ocean, Coastal and River Engineering group at the National Research Council, a team of researchers in the APEX lab, including post-doctoral research fellow Carlos Prieto, undergraduate researcher Mohammed Abusalih and Professor Schell, developed a new method for hydropower prediction using only remotely-sensed data – an improvement over time-intensive and costly field campaigns to gather in-situ flow data. Their method was recently published in Physics of Fluids, entitled “Physics-Informed Neural Network for Open Channel Flow Velocity Prediction”.