A data-driven method for fast predicting the long-term hydrodynamics of gas–solid flows: Optimized dynamic mode decomposition with control

Data-driven methods are of great interest in studying the hydrodynamics of gas–solid flows. In this paper, we developed an optimized dynamic mode decomposition with control (DMDc) method for long-term and fast prediction of one physical field with the aid of another physical field. Using the computational fluid dynamics-discrete element method (CFD-DEM) simulation results as the benchmark, the prediction ability of the standard DMDc method and the optimized DMDc method is evaluated. It was shown that the optimized DMDc method is superior when the order of magnitude of the predicted data is much larger than that of the auxiliary data, which cannot be addressed by using scaled or dimensionless data, for instance, the prediction of gas pressure with the aid of solid volume fraction; on the other hand, both DMDc and optimized DMDc methods can reasonably predict the long-term behavior of gas–solid flows, when the magnitude of the elements of the predicted field is comparative to that of the auxiliary field. This study proposes a fast and relatively accurate method for predicting the hydrodynamics of gas–solid flows with the aid of a known field.