Scholarlink

<p>The effects occurring in the cases where the acoustic waves generated by a&#xa0;pulsed source are applied to linearly elastic solids formed by a&#xa0;periodic lattice structure are examined. For a&#xa0;series of numerical experiments based on the finite element method, various types of lattice structures are modeled in the CAE Fidesys. In these lattice structures, the parameters of the undulation of the bars forming the lattice structure, the frequency of the applied pulse, and the distance at which the sound insulation level is measured, are varied. A&#xa0;number of points forming the cells of the lattice structure are also varied. The presence of frequency filtering is established for some of these variable parameters. An analysis of the relationships of variable parameters and sound insulation level is made. Using a&#xa0;series of virtual experiments, a&#xa0;model predicting the level of sound insulation is built based on a&#xa0;neural network. This algorithm is configured using the Python programming language and the scikit-learn library. The use of this algorithm allows reduce the time for calculating the sound insulation level by a&#xa0;factor of hundreds.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimization of lattice structures using neural networks and numerical simulations based on FEM

  • 期刊论文

  • 发表日期：

    2025年3月1日

查看全文
  • I. A. Podpruzhnikov,
  • A. V. Vershinin,
  • V. A. Levin,
  • K. M. Zingerman

摘要

The effects occurring in the cases where the acoustic waves generated by a pulsed source are applied to linearly elastic solids formed by a periodic lattice structure are examined. For a series of numerical experiments based on the finite element method, various types of lattice structures are modeled in the CAE Fidesys. In these lattice structures, the parameters of the undulation of the bars forming the lattice structure, the frequency of the applied pulse, and the distance at which the sound insulation level is measured, are varied. A number of points forming the cells of the lattice structure are also varied. The presence of frequency filtering is established for some of these variable parameters. An analysis of the relationships of variable parameters and sound insulation level is made. Using a series of virtual experiments, a model predicting the level of sound insulation is built based on a neural network. This algorithm is configured using the Python programming language and the scikit-learn library. The use of this algorithm allows reduce the time for calculating the sound insulation level by a factor of hundreds.