Reconstruction of 3D flow field around a building model in wind tunnel: a novel physics-informed neural network framework adopting dynamic prioritization self-adaptive loss balance strategy
Physics-informed neural networks (PINNs) have recently emerged and attracted extensive attention as an alternative approach to computational fluid dynamics (CFD) methods, which can provide competitive solutions to a variety of forward and inverse fluid problems. In this study, we reconstruct a 3D wi...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2023-12-01
|
| Series: | Engineering Applications of Computational Fluid Mechanics |
| Subjects: | |
| Online Access: | https://www.tandfonline.com/doi/10.1080/19942060.2023.2238849 |
| _version_ | 1797403888415932416 |
|---|---|
| author | En-Ze Rui Zheng-Wei Chen Yi-Qing Ni Lei Yuan Guang-Zhi Zeng |
| author_facet | En-Ze Rui Zheng-Wei Chen Yi-Qing Ni Lei Yuan Guang-Zhi Zeng |
| author_sort | En-Ze Rui |
| collection | DOAJ |
| description | Physics-informed neural networks (PINNs) have recently emerged and attracted extensive attention as an alternative approach to computational fluid dynamics (CFD) methods, which can provide competitive solutions to a variety of forward and inverse fluid problems. In this study, we reconstruct a 3D wind field around a building model in wind tunnel test with a Reynolds number of 2.4 × 104 by formulating a novel PINN framework, which is the first exploration of PINNs for building wind engineering problems. To surmount the hurdle in multi-objective optimization for PINN training, a dynamic prioritization (dp) self-adaptive loss balance strategy is proposed (termed dpPINN), which adaptively reconciles the loss terms of distinct scales to facilitate convergence in PINN training. A zero-equation turbulence model and the wind velocity data collected in near-wall regions are embedded in dpPINN training. Comparison results indicate that dpPINN predictions show good consistency with observation data, which is superior to two current PINN paradigms in prediction accuracy. Furthermore, the influence of neural network configurations, turbulence models, and the layout arrangements of training points on the dpPINN prediction is investigated. It is demonstrated that the dpPINN could be a powerful auxiliary means for airflow simulation and reconstruction in wind engineering applications. |
| first_indexed | 2024-03-09T02:46:03Z |
| format | Article |
| id | doaj.art-22f91104b46c461e93befeace699ec8a |
| institution | Directory Open Access Journal |
| issn | 1994-2060 1997-003X |
| language | English |
| last_indexed | 2024-03-09T02:46:03Z |
| publishDate | 2023-12-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | Engineering Applications of Computational Fluid Mechanics |
| spelling | doaj.art-22f91104b46c461e93befeace699ec8a2023-12-05T16:53:44ZengTaylor & Francis GroupEngineering Applications of Computational Fluid Mechanics1994-20601997-003X2023-12-0117110.1080/19942060.2023.2238849Reconstruction of 3D flow field around a building model in wind tunnel: a novel physics-informed neural network framework adopting dynamic prioritization self-adaptive loss balance strategyEn-Ze Rui0Zheng-Wei Chen1Yi-Qing Ni2Lei Yuan3Guang-Zhi Zeng4Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, People’s Republic of ChinaDepartment of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, People’s Republic of ChinaDepartment of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, People’s Republic of ChinaDepartment of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, People’s Republic of ChinaDepartment of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong, People’s Republic of ChinaPhysics-informed neural networks (PINNs) have recently emerged and attracted extensive attention as an alternative approach to computational fluid dynamics (CFD) methods, which can provide competitive solutions to a variety of forward and inverse fluid problems. In this study, we reconstruct a 3D wind field around a building model in wind tunnel test with a Reynolds number of 2.4 × 104 by formulating a novel PINN framework, which is the first exploration of PINNs for building wind engineering problems. To surmount the hurdle in multi-objective optimization for PINN training, a dynamic prioritization (dp) self-adaptive loss balance strategy is proposed (termed dpPINN), which adaptively reconciles the loss terms of distinct scales to facilitate convergence in PINN training. A zero-equation turbulence model and the wind velocity data collected in near-wall regions are embedded in dpPINN training. Comparison results indicate that dpPINN predictions show good consistency with observation data, which is superior to two current PINN paradigms in prediction accuracy. Furthermore, the influence of neural network configurations, turbulence models, and the layout arrangements of training points on the dpPINN prediction is investigated. It is demonstrated that the dpPINN could be a powerful auxiliary means for airflow simulation and reconstruction in wind engineering applications.https://www.tandfonline.com/doi/10.1080/19942060.2023.2238849BuildingsFlow fieldsReconstructionPhysics-informed neural network (PINN)Wind tunnel test |
| spellingShingle | En-Ze Rui Zheng-Wei Chen Yi-Qing Ni Lei Yuan Guang-Zhi Zeng Reconstruction of 3D flow field around a building model in wind tunnel: a novel physics-informed neural network framework adopting dynamic prioritization self-adaptive loss balance strategy Engineering Applications of Computational Fluid Mechanics Buildings Flow fields Reconstruction Physics-informed neural network (PINN) Wind tunnel test |
| title | Reconstruction of 3D flow field around a building model in wind tunnel: a novel physics-informed neural network framework adopting dynamic prioritization self-adaptive loss balance strategy |
| title_full | Reconstruction of 3D flow field around a building model in wind tunnel: a novel physics-informed neural network framework adopting dynamic prioritization self-adaptive loss balance strategy |
| title_fullStr | Reconstruction of 3D flow field around a building model in wind tunnel: a novel physics-informed neural network framework adopting dynamic prioritization self-adaptive loss balance strategy |
| title_full_unstemmed | Reconstruction of 3D flow field around a building model in wind tunnel: a novel physics-informed neural network framework adopting dynamic prioritization self-adaptive loss balance strategy |
| title_short | Reconstruction of 3D flow field around a building model in wind tunnel: a novel physics-informed neural network framework adopting dynamic prioritization self-adaptive loss balance strategy |
| title_sort | reconstruction of 3d flow field around a building model in wind tunnel a novel physics informed neural network framework adopting dynamic prioritization self adaptive loss balance strategy |
| topic | Buildings Flow fields Reconstruction Physics-informed neural network (PINN) Wind tunnel test |
| url | https://www.tandfonline.com/doi/10.1080/19942060.2023.2238849 |
| work_keys_str_mv | AT enzerui reconstructionof3dflowfieldaroundabuildingmodelinwindtunnelanovelphysicsinformedneuralnetworkframeworkadoptingdynamicprioritizationselfadaptivelossbalancestrategy AT zhengweichen reconstructionof3dflowfieldaroundabuildingmodelinwindtunnelanovelphysicsinformedneuralnetworkframeworkadoptingdynamicprioritizationselfadaptivelossbalancestrategy AT yiqingni reconstructionof3dflowfieldaroundabuildingmodelinwindtunnelanovelphysicsinformedneuralnetworkframeworkadoptingdynamicprioritizationselfadaptivelossbalancestrategy AT leiyuan reconstructionof3dflowfieldaroundabuildingmodelinwindtunnelanovelphysicsinformedneuralnetworkframeworkadoptingdynamicprioritizationselfadaptivelossbalancestrategy AT guangzhizeng reconstructionof3dflowfieldaroundabuildingmodelinwindtunnelanovelphysicsinformedneuralnetworkframeworkadoptingdynamicprioritizationselfadaptivelossbalancestrategy |