Haoyan Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Haoyan Li has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Alessandro Gambale
BuildWind SRL
3 shared publications - Alberto Procacci
Université Libre de Bruxelles, École polytechnique de Bruxelles, Aero-Thermo-Mechanics Laboratory 1 , Av. Franklin Roosevelt 50, Brussels 1050
2 shared publications - Alessandro Parente
Aero-Thermo-Mechanics Laboratory, École Polytechnique de Bruxelles, Université Libre de Bruxelles, Brussels, Belgium
2 shared publications - Gabriele Mosca
BuildWind SRL
2 shared publications - Sampath Kumar Raghunathan Srikumar
BuildWind SRL
2 shared publications - A. Lukyanov
Institute of Thermophysics SB RAS , Lavrentyev Str. 1, 630090 Novosibirsk, Russia
1 shared publication
Latest Publications
- Open Access
- 1 July 2026
- Published by Elsevier in Results in Engineering
- Vol. 32
Expert-based reduced-order modelling of urban air flow and pollutant dispersion via wind-rose-informed clustering
- 1 July 2026
- Published by Elsevier in Results in Engineering
- Vol. 32
Open AccessExpert-based reduced-order modelling of urban air flow and pollutant dispersion via wind-rose-informed clustering
- 1 September 2025
- Published by AIP Publishing in Physics of Fluids
- Vol. 37 (9)
A clustering-based domain decomposition framework for reduced-order modeling: Application to atmospheric boundary layer flows
Reduced-order models emerge as a solution to efficiently predict atmospheric boundary layer flows. However, global reduced-order models, applying dimensionality reduction to the whole domain, struggle with accuracy in domains disturbed by buildings. To address this limitation, this study proposed an unsupervised domain-decomposition approach for reduced-order models using clustering to divide the domain into building-influenced and uninfluenced subdomains. Clustering enables the localized dimensionality reduction via principal component analysis and predictive mapping using Gaussian process regression in each subdomain. Two clustering methods, K-means and vector quantization principal component analysis, were used as the comparison to individually develop domain-decomposition reduced-order models. A steady atmospheric boundary layer flow over a building array across 64 operating conditions served as the test case. The development of the reduced-order models was based on computational fluid dynamics simulations using k-ω shear stress transport model. Compared to the global reduced-order model, the clustering-based domain decomposition reduced-order models reduced root mean square error by 9.5% for the stream-wise velocity field and 18.2% for the turbulent kinetic energy field and achieved a speed-up of approximately 105 times over full-order computational fluid dynamics simulations.
- 1 September 2025
- Published by AIP Publishing in Physics of Fluids
- Vol. 37 (9)
A clustering-based domain decomposition framework for reduced-order modeling: Application to atmospheric boundary layer flows
Reduced-order models emerge as a solution to efficiently predict atmospheric boundary layer flows. However, global reduced-order models, applying dimensionality reduction to the whole domain, struggle with accuracy in domains disturbed by buildings. To address this limitation, this study proposed an unsupervised domain-decomposition approach for reduced-order models using clustering to divide the domain into building-influenced and uninfluenced subdomains. Clustering enables the localized dimensionality reduction via principal component analysis and predictive mapping using Gaussian process regression in each subdomain. Two clustering methods, K-means and vector quantization principal component analysis, were used as the comparison to individually develop domain-decomposition reduced-order models. A steady atmospheric boundary layer flow over a building array across 64 operating conditions served as the test case. The development of the reduced-order models was based on computational fluid dynamics simulations using k-ω shear stress transport model. Compared to the global reduced-order model, the clustering-based domain decomposition reduced-order models reduced root mean square error by 9.5% for the stream-wise velocity field and 18.2% for the turbulent kinetic energy field and achieved a speed-up of approximately 105 times over full-order computational fluid dynamics simulations.
- Open Access
- 1 February 2024
- Published by AIP Publishing in Physics of Fluids
- Vol. 36 (2)
Data-driven turbulence modeling for fluid flow and heat transfer in peripheral subchannels of a rod bundle
This study presents a comparison of the performance of machine learning (ML) techniques, specifically multi-dimensional gene expression programming (MGEP), tensor basis neural network (TBNN), and also proposes a novel universally interpretable machine learning architecture to model the turbulent scalar flux (UIML-s) to enhance turbulence models for fluid flows at different Prandtl numbers in channels with complex shapes of walls in the channel cross section. In particular, peripheral subchannels of rod bundles are of primary interest. However, the accuracy of mean velocity and scalar distributions predicted by commonly used turbulence models still poses a challenge compared to data extracted from high-fidelity eddy-resolving numerical simulations, particularly for engineering applications involving complex geometry flows. In the present study, by utilizing an explicit algebraic expression for the nonlinear Reynolds-stress term obtained through both the evolutionary MGEP optimization and TBNN, the secondary flow structure has been adequately predicted in the cross-wise mean velocity distributions in the square duct and the rectangular channel with three longitudinal rods. This structure is also observed in the data from the concurrent runs performed by direct numerical simulation (DNS) but is completely absent in the results produced by a baseline Reynolds-averaged Navier–Stokes (RANS) closure, which employs the linear eddy viscosity model for the Reynolds stress tensor. Comparison of MGEP and TBNN has shown their nearly equal performance in a square duct flow; however, MGEP works better for the more complex geometry channel with three rods. Furthermore, based on the velocity field produced by the RANS-MGEP model, the ML modification of the gradient diffusion hypothesis, integrated into the aforementioned novel RANS-ML model called as UIML-s, significantly improves the mean scalar distributions in a flow with three bumps serving as a prototype for the peripheral subchannel of rod bundle. The normalized root mean squared error decreases from 13.5% to 7.6%, bringing the predicted distributions closer to the DNS data, particularly in the near-wall region. Another approach, MGEP-s, also yields the acceptable results, which are nearly identical to those from UIML-s. These findings highlight the potential of using data-driven calibration of turbulence models with nonlinear closures to enhance the predictability for RANS simulations of fluid flows, heat, and mass transfer in channels with complex geometry.
- 1 February 2024
- Published by AIP Publishing in Physics of Fluids
- Vol. 36 (2)
Open AccessData-driven turbulence modeling for fluid flow and heat transfer in peripheral subchannels of a rod bundle
This study presents a comparison of the performance of machine learning (ML) techniques, specifically multi-dimensional gene expression programming (MGEP), tensor basis neural network (TBNN), and also proposes a novel universally interpretable machine learning architecture to model the turbulent scalar flux (UIML-s) to enhance turbulence models for fluid flows at different Prandtl numbers in channels with complex shapes of walls in the channel cross section. In particular, peripheral subchannels of rod bundles are of primary interest. However, the accuracy of mean velocity and scalar distributions predicted by commonly used turbulence models still poses a challenge compared to data extracted from high-fidelity eddy-resolving numerical simulations, particularly for engineering applications involving complex geometry flows. In the present study, by utilizing an explicit algebraic expression for the nonlinear Reynolds-stress term obtained through both the evolutionary MGEP optimization and TBNN, the secondary flow structure has been adequately predicted in the cross-wise mean velocity distributions in the square duct and the rectangular channel with three longitudinal rods. This structure is also observed in the data from the concurrent runs performed by direct numerical simulation (DNS) but is completely absent in the results produced by a baseline Reynolds-averaged Navier–Stokes (RANS) closure, which employs the linear eddy viscosity model for the Reynolds stress tensor. Comparison of MGEP and TBNN has shown their nearly equal performance in a square duct flow; however, MGEP works better for the more complex geometry channel with three rods. Furthermore, based on the velocity field produced by the RANS-MGEP model, the ML modification of the gradient diffusion hypothesis, integrated into the aforementioned novel RANS-ML model called as UIML-s, significantly improves the mean scalar distributions in a flow with three bumps serving as a prototype for the peripheral subchannel of rod bundle. The normalized root mean squared error decreases from 13.5% to 7.6%, bringing the predicted distributions closer to the DNS data, particularly in the near-wall region. Another approach, MGEP-s, also yields the acceptable results, which are nearly identical to those from UIML-s. These findings highlight the potential of using data-driven calibration of turbulence models with nonlinear closures to enhance the predictability for RANS simulations of fluid flows, heat, and mass transfer in channels with complex geometry.