Scholarlink

<p>In the design and analysis of complex mechanical structures, various forms of hybrid uncertainties are commonly encountered. Their propagation analysis is critical for ensuring structural design reliability. To enhance the efficiency and accuracy of hybrid uncertainty propagation, this paper proposes a rapid analysis method based on low-rank approximation (LRA). First, for structural uncertainty parameters represented in various forms, a targeted sequential sampling method is employed to effectively balance the uncertainty distribution of parameters and the local approximation accuracy of the model. Subsequently, the LRA method constructs a data-driven model that maps hybrid uncertainty parameters to the corresponding structural responses, significantly improving the efficiency and accuracy of the propagation analysis. Finally, parameterized probability boxes are used as the uncertainty quantification method for the response results based on the LRA model. Several numerical examples and engineering cases demonstrate the efficiency and the accuracy of the proposed propagation approach.</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

Efficient hybrid uncertainty propagation in mechanical structures via low-rank approximation

  • 期刊论文

  • 发表日期：

    2026年4月9日

查看全文
  • Yue Zhao,
  • Jinning Zhi,
  • Zhibin Yao

摘要

In the design and analysis of complex mechanical structures, various forms of hybrid uncertainties are commonly encountered. Their propagation analysis is critical for ensuring structural design reliability. To enhance the efficiency and accuracy of hybrid uncertainty propagation, this paper proposes a rapid analysis method based on low-rank approximation (LRA). First, for structural uncertainty parameters represented in various forms, a targeted sequential sampling method is employed to effectively balance the uncertainty distribution of parameters and the local approximation accuracy of the model. Subsequently, the LRA method constructs a data-driven model that maps hybrid uncertainty parameters to the corresponding structural responses, significantly improving the efficiency and accuracy of the propagation analysis. Finally, parameterized probability boxes are used as the uncertainty quantification method for the response results based on the LRA model. Several numerical examples and engineering cases demonstrate the efficiency and the accuracy of the proposed propagation approach.