Numerical design methods with polymorphic uncertainty modelling

具有多态不确定性建模的数值设计方法

• 批准号: 276783706
• 负责人: Professor Dr.-Ing. Wolfgang Graf
• 金额: --
• 依托单位: Institut für Statik und Dynamik der Tragwerke
• 依托单位国家: 德国
• 项目类别: Research Grants
• 财政年份: 2015
• 资助国家: 德国
• 起止时间: 2014-12-31 至 2018-12-31
• 项目状态: 已结题
• 来源: https://gepris.dfg.de/gepris/projekt/276783706?language=en
• 关键词: Numerical; design; methods; polymorphic; uncertainty

With the proposed research project, a paradigm shift is introduced in the numerical design of structures. The current state is characterized by deterministic thinking and practice. The representation of reality with deterministic models suggests accuracy and precision. Reality is the fact that all available data are characterized by uncertainty, incompleteness and inaccuracy (polymorphic uncertainty) which have to be taken into account appropriately and cannot be neglected.In the research project, innovative approaches and practices for structural design are developed that take the uncertainty of the data in all phases of design into account and to process it by numerical models to gain a new design quality. Engineering-oriented tools for the numerical design of structures under static and/or dynamic loads with consequent consideration of polymorphic data uncertainty need to be developed.The applied numerical simulations come along with high costs. This numerical effort increases exponentially, if the uncertainty of data is taken into account consistently and correctly. In addition to the use of appropriate uncertainty models, capturing the spatial and time dependency is required. On the basis of (relatively few) realistic, deterministic, nonlinear, numerical simulations, metamodels for the approximation of relevant result variables and sensitivity analysis have to be developed. The application potential for a numerical design of structures with polymorphic uncertainty modeling is very large and not limited to the Civil Engineering.

通过提出的研究项目，在结构的数值设计中引入了一种范式转换。当前状态的特点是确定性思维和实践。用确定性模型来表示现实意味着准确性和精确性。现实情况是，所有可用的数据都具有不确定性、不完整性和不准确性（多态不确定性）的特点，必须适当地考虑到这一点，不能忽视。在研究项目中，开发了结构设计的创新方法和实践，将设计各个阶段数据的不确定性考虑在内，并通过数值模型对其进行处理，以获得新的设计质量。需要开发面向工程的工具，用于静态和/或动态载荷下的结构数值设计，从而考虑多态数据的不确定性。应用数值模拟的成本很高。如果始终正确地考虑数据的不确定性，这种数值努力将呈指数级增长。除了使用适当的不确定性模型外，还需要捕获空间和时间依赖性。在（相对较少的）现实的、确定性的、非线性的、数值模拟的基础上，必须开发用于近似相关结果变量和敏感性分析的元模型。多态不确定性模型在结构数值设计中的应用潜力非常大，不仅限于土木工程。

项目成果

Data Mining and Machine Learning Methods Applied to A Numerical Clinching Model

• DOI: 10.31614/cmes.2018.04112
• 发表时间: 2018-12
• 期刊: Computer Modeling in Engineering & Sciences
• 作者: Marco Götz;F. Leichsenring;T. Kropp;P. Müller;T. Falk;W. Graf;M. Kaliske;W. Drossel

Structural design with polymorphic uncertainty models

具有多态不确定性模型的结构设计

• DOI: 10.1504/ijrs.2015.072715
• 发表时间: 2015
• 期刊: International Journal of Reliability and Safety
• 作者: Kaliske

Computing permissible design spaces under consideration of functional responses

• DOI: 10.1016/j.advengsoft.2017.05.015
• 发表时间: 2017-06
• 期刊: Adv. Eng. Softw.
• 作者: W. Graf;Marco Götz;M. Kaliske

Analysis of dynamical processes under consideration of polymorphic uncertainty

• DOI: 10.1016/j.strusafe.2014.09.003
• 发表时间: 2015
• 期刊: Structural Safety
• 影响因子: 5.8
• 作者: W. Graf;Marco Götz;M. Kaliske

Assessment and design of an engineering structure with polymorphic uncertainty quantification

• DOI: 10.1002/gamm.201900009
• 发表时间: 2019-04
• 期刊: GAMM‐Mitteilungen
• 作者: I. Papaioannou;M. Daub;Martin Drieschner;F. Duddeck;Max Ehre;L. Eichner;M. Eigel;Marco Götz;W. Graf;L. Grasedyck;Robert Gruhlke;D. Hömberg;M. Kaliske;Dieter Moser;Y. Petryna;D. Štraub