Skip to main content

Physics-informed deep learning and field experiments for dynamic reactive fluids (carbon, nutrients and water flow) in agricultural soils

Academic lead
Xiaohui Chen​, Civil Engineering
Industrial lead
Nick Humphries, ​Anglo American​
Co-supervisor(s)
Kenny Brown, N2-Applied​, Shashank Bettadapura Subramanyam​, Civil Engineering​ , Laura Carter, Geography
Project themes
Environmental Flows, Industrial Processes, Underpinning Methods for Fluid Dynamics

Artificial Intelligence have attracted considerable attention from researchers in environmental fluid dynamics over the last decade, especially Artificial Neural Networks (ANN) which can provide a flexible mathematical structure capable of identifying complex nonlinear relationships between input and output data sets. However, traditionally, ANNs have been trained using input and output datasets with simple loss functions, which incorporates no physical system knowledge into the learning process, while requiring huge amounts of data, which are either costly to produce or unavailable. Alternatively, conceptual and computational modelling has continued rapid development in the past decades as it is based on fundamental constitutive physical equations and able to provide accurate analysis and prediction, however, it suffers from problems associated with computational performance, which can hinder its usage. This project will integrate fundamental physics into the training process of deep learning processes, engaging computational modelling and deep neural network, and establish a new generation of physics informed deep learning methods, which can provide accurate and quick estimates of environmental fluids in soil system response. This project will be using a real field plot at Leeds University Farm to provide date and training for the PINN.