Imagen de Portada

A flow pattern recognition method for gas-liquid two-phase flow based on dilated convolutional channel attention mechanism

Guardado en:
Publicado en:PLoS One vol. 20, no. 6 (Jun 2025), p. e0325784
Autor principal: Liu, Jie
Otros Autores: Wu, Yang
Publicado:
Public Library of Science
Materias:
Flow distribution
Feature extraction
Pattern recognition
Flow pattern
Datasets
Data processing
Nuclear reactors
Identification methods
Neural networks
Signal processing
Two phase flow
Information processing
Slug flow
Deep learning
Machine learning
Multiphase flow
Transfer learning
Transition layers
Environmental
Acceso en línea:Citation/Abstract
Full Text
Full Text - PDF
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
Resumen:Addressing the issue of insufficient key feature extraction leading to low recognition rates in existing deep learning-based flow pattern identification methods, this paper proposes a novel flow pattern image recognition model, Enhanced DenseNet with transfer learning (ED-DenseNet). The model enhances the deep feature extraction capability by introducing a multi-branch structure, incorporating an ECA attention mechanism into Dense Blocks and dilated convolutions into Transition Layers to achieve multi-scale feature extraction and refined channel information processing. Considering the limited scale of the experimental dataset, pretrained DenseNet121 weights on ImageNet were transferred to ED-DenseNet using transfer learning. On a gas-liquid two-phase flow image dataset containing Annular, Bubbly, Churn, Dispersed, and Slug flow patterns, ED-DenseNet achieved an overall recognition accuracy of 97.82%, outperforming state-of-the-art models such as Flow-Hilbert–CNN, especially in complex and transitional flow scenarios. Additionally, the model’s generalization and robustness were further validated on a nitrogen condensation two-phase flow dataset, demonstrating superior adaptability compared to other methods.
ISSN:1932-6203
DOI:10.1371/journal.pone.0325784
Fuente:Health & Medical Collection