Abstract

To examine the design direction of side mirrors in the early stages of development, a Pix2Pix GAN deep learning model was used to develop an easy-to-operate tool that enables prediction of flow fields with sufficient accuracy in a short period of time and output of the results. In addition to a shorter prediction time with a deep learning model, further time reduction was achieved by omitting the signed distance function for shape recognition that has a high computational load. The tool was also developed to maintain prediction accuracy even with a small amount of training data. Furthermore, output of fluid physical quantities required for verification on the tool operation screen was minimized, allowing performance of the process from input to output solely by button operations, making it easy to use for designers.

References

1. (1) Akasaka, K., Chen, F., Teraguchi, T.: Surrogate Model Development for Prediction of Car Aerodynamics Using Machine Learning, Transactions of Society of Automotive Engineers of Japan, Vol. 52, Issue 3, p. 621-626, (2021) (in Japanese), doi: 10.11351/jsaeronbun.52.621
2. (2) Guo, X., Li, W., Iorio, F.: Convolutional Neural Networks for Steady Flow Approximation, KDD ’16: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, p. 481-490, (2016), doi: 10.1145/2939672.2939738
3. (3) Isola, P., Zhu, J-Y., Zhou, T., Efros, A. A.: Image-to-Image Translation with Conditional Adversarial Networks, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), p. 5967-5976, (2017), doi: 10.1109/CVPR.2017.632
4. (4) Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., Bengio, Y.: Generative adversarial networks, Communications of the ACM, Vol. 63, Issue 11, p. 139-144, (2020), doi: 10.1145/3422622
5. (5) Mirza, M., Osindero, S.: Conditional Generative Adversarial Nets, arXiv, (2014), doi: 10.48550/arXiv.1411.1784
6. (6) Cirillo, M. D., Abramian, D., Eklund, A.: Vox2Vox: 3D-GAN for Brain Tumour Segmentation, Lecture Notes in Computer Science, Springer, Cham, Vol. 12658, p. 274-284, (2021), doi: 10.1007/978-3-030-72084-1_25
7. (7) Ronneberger, O., Fischer, P., Brox, T.: U-Net: Convolutional Networks for Biomedical Image Segmentation, Lecture Notes in Computer Science, Vol. 9351, p. 234-241, (2015), doi: 10.1007/978-3-319-24574-4_28

Author (organization or company)

• Takuro HASHIMOTO（Honda Motor Co., Ltd.）
• Koji OKABE（Honda Motor Co., Ltd.）
• Takeshi YAMATODA（IDAJ Co., LTD.）
• Takafusa KIDO（IDAJ Co., LTD.）

Publication Information

Article of Honda R&D Technical Review Vol.38 (123 - 130)

PDF

Citation Information