First Running! Deep learning to CFD

This is our first apply ConvLSTM to CFD successfully! although the case is simple and under control of lots of factors. The ground factor is generated by Openfoam, and the custom model is predNet from coxlab. We trained three models in this time.

1.Training: use Nth frame to predict (N+1)th frame
Prediction: use 1-10th frame to predict 2-11th frame, then combined 11th frame in the predicted output with 1-10th frame. With new input(2-10th are ground truth, 11th is predicted), we can keep predicting 3-12th frame. In this experience, we predict the frames until 20th where sliding window = 1 frame. ( only first few frame are good, since we use the predicting frames to do the prediction)

LSTM

2.Training: Nth frame to predict (N+10)th frame
Prediction: use 1-10 frame to predict 11-20 frame, no sliding window. ( this is very good since all input are ground truth)

LSTM

3.Training: use Nth frame to predict (N+1)th frame
Prediction: use one frame to predict next frame, like driving prediction (animation has a little problem. right side is prediction, left side is ground truth)

LSTM

  1. Sandipan Banerjee says:

    Hey,

    Great work, I am trying to do something similar, and want to ask you, the following questions:

    1. Are your inputs in images or are they as physical quantities like the velocity field, pressure, etc?
    2. If the inputs are physical quantities, then for each grid point you have how many features, and would it be possible to provide a sample of your code specially the section where you include the data in the spatial and temporal format?

Leave a Reply