航空学报 > 2023, Vol. 44 Issue (22): 28611-028611   doi: 10.7527/S1000-6893.2023.28611

基于数值模拟的工程设计中参数不确定性表征方法研究综述

熊芬芬1(), 李泽贤1, 刘宇2, 夏侯唐凡2   

  1. 1.北京理工大学 宇航学院,北京 100081
    2.电子科技大学 机械与电气工程学院,成都 611731
  • 收稿日期:2023-02-24 修回日期:2023-03-22 接受日期:2023-04-21 出版日期:2023-11-25 发布日期:2023-04-21
  • 通讯作者: 熊芬芬 E-mail:fenfenx@bit.edu.cn
  • 基金资助:
    国家自然科学基金(52175214);基础研究项目群(514010103-302)

A review of characterization methods for parameter uncertainty in engineering design based on numerical simulation

Fenfen XIONG1(), Zexian LI1, Yu LIU2, Tangfan XIAHOU2   

  1. 1.School of Aerospace Engineering,Beijing Institute of Technology,Beijing 100081,China
    2.School of Mechanical and Electrical Engineering,University of Electronic Science and Technology of China,Chengdu 611731,China
  • Received:2023-02-24 Revised:2023-03-22 Accepted:2023-04-21 Online:2023-11-25 Published:2023-04-21
  • Contact: Fenfen XIONG E-mail:fenfenx@bit.edu.cn
  • Supported by:
    National Natural Science Foundation of China(52175214);The Basic Research Program(514010103-302)

摘要:

现代工程设计大量使用数值模拟技术,然而数值模拟中仿真模型、几何参数、工作载荷、环境条件等均存在不确定性,这些不确定性直接影响数值模拟结果的可信度,尤其在研发对性能和可靠性要求严苛的装备时,忽略这些不确定性将产生潜在风险,因此在基于数值模拟的工程设计中开展不确定性量化意义重大。不确定性表征既是准确开展不确定性量化和优化设计工作的前提,也是工程精细化设计的重要支撑。本文阐述了基于数值模拟的工程设计中所面临的不确定性因素,根据不确定性因素的表现形式和可用信息,按照概率表征和非概率表征两大类总结了目前主流的参数不确定性表征方法,介绍了各种表征方法的基本思想、主要原理、适用范围以及方法在工程实践中的发展和应用,并简要阐述基于不确定性表征结果进行不确定性传播的基本思路,最后对不确定性表征的研究方向进行了展望。

关键词: 数值模拟, 不确定性表征, 参数不确定性, 随机不确定性, 认知不确定性

Abstract:

Numerical simulation technology is widely used in modern engineering design. However, uncertainties are ubiquitous in simulation models: geometric parameters, working loads, environmental conditions, etc., which directly affect the credibility of numerical simulation results. Especially for equipment with strict requirements on performance and reliability, ignoring these uncertainties will cause potential risks. Therefore, it is of great significance to carry out uncertainty quantification in engineering design based on numerical simulation. Uncertainty characterization is not only a prerequisite for accurate uncertainty quantification and optimal design but also an important support for engineering refinement design. This paper elucidates the uncertainty factors in engineering design based on numerical simulation. Considering the manifestation and available information of the uncertainty factors, the current mainstream characterization methods of parameter uncertainty according to the two categories of probabilistic and non-probabilistic characterization methods are summarized. The basic ideas, main principles, application scope, and development and application of various characterization methods in engineering practice are introduced. The basic ideas of uncertainty propagation based on uncertainty characterization results are also briefly explained. Finally, the research direction of uncertainty characterization is prospected.

Key words: numerical simulation, uncertainty characterization, parameter uncertainty, aleatory uncertainty, epistemic uncertainty

中图分类号: