A generalised CFD learning and prediction system

An expert system that learns data from Computational Fluid Dynamics (CFD) simulations and presents it to non-expert users of CFDs to make informed design decisions is proposed. The system is responsible for choosing CFDs to run, supervising their running, determining when to halt them and then teaching them to a Neural Network (NN). The system also includes a visualisation tool which displays the effect of CFDs on the output of the NN. A case study is finally presented to demonstrate how the tool has been used with fire simulations.