基于DNN的离心螺旋流场湍流耗散建模
DOI:
作者:
作者单位:

广西科技大学机械与汽车工程学院

作者简介:

通讯作者:

中图分类号:

O357.5

基金项目:


Deep neural network-based centrifugal device spiral Turbulent Dissipation Study of Flow Field
Author:
Affiliation:

School of Mechanical and Automotive Engineering, Guangxi University of Science and Technology

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    传统的湍流问题求解方法需要耗费大量的计算资源,而基于机器学习的模型在湍流建模领域展现出潜在的前景,但目前主要局限于特定流场。为此,本文提出了一种针对螺旋流场的基于深度学习的湍动能耗散率预测模型。首先,通过对离心风机流场进行数值模拟来获取数据集,然后根据湍动能耗散理论模型和运行工况选择输入特征,使用神经网络算法建立耗散率预测模型并进行训练,最后对未训练区域和不同转速下的离心风机流场进行模型测试,并将其应用于卧螺离心机流场和通道流场。结果表明,所构建的模型在离心风机流场的不同区域和转速下均表现出较好的耗散率预测能力,决定系数(R2)均在0.9以上,并且能够准确预测卧螺离心机流场和通道流场湍流耗散的变化趋势。这一研究为解决复杂流场湍流问题提供了新的思路,同时也为机器学习在流场研究中的应用提供了借鉴。

    Abstract:

    While traditional methods for solving turbulence problems are computationally intensive, machine learning-based models show potential promise in the field of turbulence modeling, but are currently mainly limited to specific flow fields. To this end, a deep learning-based turbulent kinetic energy dissipation rate prediction model for helical flow fields is proposed in this paper. Firstly, the dataset is obtained by numerical simulation of centrifugal fan flow field, then the input features are selected according to the theoretical model of turbulent kinetic energy dissipation and operating conditions, the dissipation rate prediction model is built and trained using the neural network algorithm, and finally, the model is tested for the centrifugal fan flow field in the untrained region and at different rotational speeds, and it is applied to the flow field of the decanter centrifuge and the channel flow field. The results show that the constructed models exhibit good dissipation rate prediction ability in different regions and speeds of centrifugal fan flow fields, with the coefficient of determination (R2) above 0.9, and can accurately predict the trend of turbulent dissipation in the decanter centrifuge flow field and the channel flow field. This study provides new ideas for solving complex flow field turbulence problems, and also provides a reference for the application of machine learning in flow field research.

    参考文献
    相似文献
    引证文献
引用本文

丛佩超. 基于DNN的离心螺旋流场湍流耗散建模[J]. 科学技术与工程, , ():

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2024-04-05
  • 最后修改日期:2024-06-06
  • 录用日期:2024-07-09
  • 在线发布日期:
  • 出版日期:
×
亟待确认版面费归属稿件,敬请作者关注
《科学技术与工程》入选维普《中文科技期刊数据库》自然科学类期刊月度下载排行榜TOP10