Scholarlink

<p>Aiming at the problem that the calculation of solid finite element model is slow (average calculation time approximately 2137 s) and the error of traditional simplified beam element model is large (15–27%) in the iterative design and certification of landing gear, this study proposes an automatic correction method for section scaling coefficients based on a deep neural network. By constructing a “stiffness section coefficient” dataset, the parameters of the beam element model can be inversely solved. The results show that the section prediction error is less than 5%, the global stiffness error of the modified beam model is less than 1%, and the simulation response error is less than 10%, which meets the engineering requirements of the preliminary design stage. This method shortens the correction time to within 120 s, reduces the dependence on manual tuning, and provides a feasible path for rapid correction of landing gear stiffness in engineering applications.</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

A Stiffness Inversion and Correction Method for Aircraft Landing Gear Beam Element Model Based on Deep Learning

  • 期刊论文

  • 发表日期：

    2026年4月9日

查看全文
  • Bohao Fang,
  • Ran Guo,
  • Ji Yao,
  • Chang Liu,
  • Runjie Wang,
  • Xuran Li

摘要

Aiming at the problem that the calculation of solid finite element model is slow (average calculation time approximately 2137 s) and the error of traditional simplified beam element model is large (15–27%) in the iterative design and certification of landing gear, this study proposes an automatic correction method for section scaling coefficients based on a deep neural network. By constructing a “stiffness section coefficient” dataset, the parameters of the beam element model can be inversely solved. The results show that the section prediction error is less than 5%, the global stiffness error of the modified beam model is less than 1%, and the simulation response error is less than 10%, which meets the engineering requirements of the preliminary design stage. This method shortens the correction time to within 120 s, reduces the dependence on manual tuning, and provides a feasible path for rapid correction of landing gear stiffness in engineering applications.