Verification of Visualization Method Using Self-Organizing Map for Various Flow Fields

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

Please login to view abstract download link

The present research is an ongoing study to propose a new visualization method using machine learning techniques to simplify the visualization process. In the previous studies, we proposed a visualization method using Self-Organizing Map (SOM) for two-dimensional flows. Flow variables, namely velocity, pressure and their spatial gradients, of each grid point are classified by SOM and the color is given according to the location on the map so as to paint the location of the same flow feature with the same color. In this conference, we discuss the effectiveness of the present method, comparing with the existing methods as well as a yet another, but deterministic, method.