Prediction of CFD analysis results using deep learning

  • Masuda, Masato (Toyo University)
  • Nakabayashi, Yasushi (Toyo University)
  • Tamura, Yoshiaki (Toyo University)

Please login to view abstract download link

In this study, we propose a technique for predicting numerical fluid analysis results in the next step using information from several previous analysis steps in a Cartesian grid. Already we have successfully predicted the analysis results with visualized data. In this study, we will examine the prediction accuracy and validity of the prediction by giving the analytical results as physical quantities to the input data.