An attention-enhanced Fourier neural operator model for predicting flow fields in turbomachinery Cascades DOI

Lele Li,

Weihao Zhang, Ya Li

et al.

Physics of Fluids, Journal Year: 2025, Volume and Issue: 37(3)

Published: March 1, 2025

Flow field information within cascades is crucial for refined turbomachinery design. Currently, this primarily obtained through experimental methods or numerical simulations, both of which are complex and time-consuming. Data-driven deep learning approaches offer a potential solution rapid flow evaluation. However, existing learning-based prediction models exhibit certain limitations in accuracy generalization, particularly regions with high gradients, often the primary sources aerodynamic losses. To address these issues, study develops high-precision cascade model, A-FNO, based on Galerkin-type self-attention mechanism Fourier Neural Operator (FNO). A-FNO designed newly proposed FNO, has demonstrated excellent performance solving partial differential equations. This extends its application to problems. mitigate FNO predicting areas steep gradient changes, we incorporate capture dependencies between different field, thereby enhancing FNO's ability express details. Experimental results demonstrate that significantly improves surrounding boundary layer. The maximum relative error velocity predictions 5%, pressure 2%, temperature 1%.

Language: Английский

Weakly-supervised thyroid ultrasound segmentation: Leveraging multi-scale consistency, contextual features, and bounding box supervision for accurate target delineation DOI
Mohammed Aly

Computers in Biology and Medicine, Journal Year: 2025, Volume and Issue: 186, P. 109669 - 109669

Published: Jan. 13, 2025

Language: Английский

Citations

0
PR‐YOLOv9: An improve defect detection network for hot‐pressed light guide plates DOI Open Access

Cunling Liu,

Shuo Peng,

Shuangning Liu

et al.

Journal of the Society for Information Display, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 6, 2025

Abstract As one of the key components liquid crystal display, quality hot‐pressed light guide plate (LGP) directly affects display performance. To address challenges posed by complex background textures, diverse types defects, large variations in defect resolutions, and low contrast, this paper proposes a surface detection method for LGPs based on PR‐YOLOv9. The poly kernel inception network (PKINet) module is integrated replacing second convolution YOLOv9 backbone network, effectively reducing interference from invalid targets such as textured backgrounds, thereby enhancing network's ability to detect multi‐scale defects decreasing parameters. Additionally, receptive‐field attention convolutional operation (RFAConv) incorporated, first last layers with module. RFAConv provides weights kernels, improving extract spatial feature information. Experimental results show that proposed PR‐YOLOv9 achieves mean average precision (mAP) 98.40% F1‐Score 97.14% self‐constructed LGP dataset, reduction 6.19 M parameters compared YOLOv9, representing decrease 10.18%, making it suitable real‐time industrial settings.

Language: Английский

Citations

0
An attention-enhanced Fourier neural operator model for predicting flow fields in turbomachinery Cascades DOI

Lele Li,

Weihao Zhang, Ya Li

et al.

Physics of Fluids, Journal Year: 2025, Volume and Issue: 37(3)

Published: March 1, 2025

Flow field information within cascades is crucial for refined turbomachinery design. Currently, this primarily obtained through experimental methods or numerical simulations, both of which are complex and time-consuming. Data-driven deep learning approaches offer a potential solution rapid flow evaluation. However, existing learning-based prediction models exhibit certain limitations in accuracy generalization, particularly regions with high gradients, often the primary sources aerodynamic losses. To address these issues, study develops high-precision cascade model, A-FNO, based on Galerkin-type self-attention mechanism Fourier Neural Operator (FNO). A-FNO designed newly proposed FNO, has demonstrated excellent performance solving partial differential equations. This extends its application to problems. mitigate FNO predicting areas steep gradient changes, we incorporate capture dependencies between different field, thereby enhancing FNO's ability express details. Experimental results demonstrate that significantly improves surrounding boundary layer. The maximum relative error velocity predictions 5%, pressure 2%, temperature 1%.

Language: Английский

Citations

0