한국가시화정보학회지 (Journal of the Korean Society of Visualization)

  • 제21권1호
  • /
  • Pages.119-126
  • /
  • 2023
  • /
  • 1598-8430(pISSN)
  • /
  • 2093-808X(eISSN)
한국가시화정보학회 (The Korean Society of Visualization)
권호 표지
DOI QR코드

DOI QR Code

물리 기반 인공신경망을 이용한 PIV용 합성 입자이미지 생성

Generation of Synthetic Particle Images for Particle Image Velocimetry using Physics-Informed Neural Network

  • Hyeon Jo Choi (School of Mechanical Engineering, Chonnam National University) ;
  • Myeong Hyeon, Shin (School of Mechanical Engineering, Chonnam National University) ;
  • Jong Ho, Park (School of Mechanical Engineering, Chonnam National University) ;
  • Jinsoo Park (School of Mechanical Engineering, Chonnam National University)
  • 투고 : 2023.02.24
  • 심사 : 2023.03.28
  • 발행 : 2023.03.31
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

Acquiring experimental data for PIV verification or machine learning training data is resource-demanding, leading to an increasing interest in synthetic particle images as simulation data. Conventional synthetic particle image generation algorithms do not follow physical laws, and the use of CFD is time-consuming and requires computing resources. In this study, we propose a new method for synthetic particle image generation, based on a Physics-Informed Neural Networks(PINN). The PINN is utilized to infer the flow fields, enabling the generation of synthetic particle images that follow physical laws with reduced computation time and have no constraints on spatial resolution compared to CFD. The proposed method is expected to contribute to the verification of PIV algorithms.

키워드

과제정보

이 연구는 전남대학교 교내연구비, 정부(산업통상자원부)의 재원으로 한국산업기술진흥원의 지원(P0011931, 광주 빛그린 산학융합지구 조성사업)을 받아 수행된 연구결과입니다.

참고문헌

  1. T. Liu and D. M. Salazar, "OpenOpticalFlow_PIV: An Open Source Program Integrating Optical Flow Method with Cross- Correlation Method for Particle Image Velocimetry," Journal of Open Research Software, vol. 9, no. 1. Ubiquity Press, Ltd., p. 3, 05-Feb-2021.
  2. J. H. Seong, M. S. Song, D. Nunez, A. Manera, and E. S. Kim, "Velocity refinement of PIV using global optical flow," Experiments in Fluids, vol. 60, no. 11. Springer Science and Business Media LLC, 29-Oct-2019.
  3. D. I. Zaripov, R. Li, N. I. Mikheev, and N. S. Dushin, "Speed-up algorithm based on parallel projection correlation technique for planar PIV: Accuracy and limitation," Flow Measurement and Instrumentation, vol. 60. Elsevier BV, pp. 88-94, Apr-2018. https://doi.org/10.1016/j.flowmeasinst.2018.02.019
  4. J. Park, S. Im, H. J. Sung, and J. S. Park, "PIV measurements of flow around an arbitrarily moving free surface," Experiments in Fluids, vol. 56, no. 3. Springer Science and Business Media LLC, Mar-2015.
  5. S. Cai, J. Liang, Q. Gao, C. Xu, and R. Wei, "Particle Image Velocimetry Based on a Deep Learning Motion Estimator," IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 6. Institute of Electrical and Electronics Engineers (IEEE), pp. 3538-3554, Jun-2020. https://doi.org/10.1109/TIM.2019.2932649
  6. Y. Lee, H. Yang, and Z. Yin, "PIV-DCNN: cascaded deep convolutional neural networks for particle image velocimetry," Experiments in Fluids, vol. 58, no. 12. Springer Science and Business Media LLC, 15-Nov-2017.
  7. L. Mendes, A. Bernardino, and R. M. L. Ferreira, "piv-image-generator: An image generating software package for planar PIV and Optical Flow benchmarking," SoftwareX, vol. 12. Elsevier BV, p. 100537, Jul-2020.
  8. M. Raissi, P. Perdikaris, and G. E. Karniadakis, "Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations," Journal of Computational Physics, vol. 378. Elsevier BV, pp. 686-707, Feb-2019. https://doi.org/10.1016/j.jcp.2018.10.045
  9. L. Sun, H. Gao, S. Pan, and J.-X. Wang, "Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data," Computer Methods in Applied Mechanics and Engineering, vol. 361. Elsevier BV, p. 112732, Apr-2020.
  10. C. Rao, H. Sun, and Y. Liu, "Physics-informed deep learning for incompressible laminar flows," Theoretical and Applied Mechanics Letters, vol. 10, no. 3. Elsevier BV, pp. 207-212, Mar-2020. https://doi.org/10.1016/j.taml.2020.01.039
  11. Raffell M, Willert C, Scarano F, Kahler C, Wereley S, Kompenhans J. "Particle image velocimetry - a practical guide", Springer, 2018, pp. 216-218, pp. 156
  12. J. Taylor et al., "Optimizing the optimizer for data driven deep neural networks and physics informed neural networks," arXiv preprint arXiv:2205.07430, 2022.
  13. U. Ghia, K. Ghia, and C. Shin, "High-Re solutions for incompressible flow using the Navier-Stokes equations and a multigrid method," Journal of Computational Physics, vol. 48, no. 3. Elsevier BV, pp. 387-411, Dec-1982.  https://doi.org/10.1016/0021-9991(82)90058-4