Accuracy Improvement Technique of DNN for Accelerating CFD Simulator

TL;DR

This paper introduces two novel techniques to enhance the accuracy and speed of DNN-based CFD simulations, achieving up to 3.9x overall acceleration while maintaining result quality.

Contribution

The paper proposes two specific methods for improving DNN inference in CFD, including a steady-state flow prediction technique and a customized loss function for better accuracy.

Findings

First technique achieves 1.7x speedup with maintained accuracy.

Second technique further improves speed to 3.9x combined.

Methods effectively reduce CFD computational costs.

Abstract

There is a Computational fluid dynamics (CFD) method of incorporating the DNN inference to reduce the computational cost. The reduction is realized by replacing some calculations by DNN inference. The cost reduction depends on the implementation method of the DNN and the accuracy of the DNN inference. Thus, we propose two techniques suitable to infer flow field on the CFD grid. The first technique is to infer the flow field of the steady state from the airfoil shape. We use the position on the coordinates of the grid point and the distance from the surface of the airfoil as input information for the DNN. The second method uses the customized mean square error as a loss function. The size of the associated area for each grid point was multiplied by the square error. This method compensates for the effect caused by the size of the associated area of nonuniform allocation of grid points. The evaluation results show that the CFD incorporated first technique achieves 1.7x speedup against the CFD without DNN, while maintaining equivalent result quality. By implementing the second technique, the CFD realized further 2.3x speed up against the CFD with first technique only.