当前位置:
X-MOL 学术
›
Comput. Methods Appl. Mech. Eng.
›
论文详情
Our official English website, www.x-mol.net, welcomes your
feedback! (Note: you will need to create a separate account there.)
An approximate decoupled reliability-based design optimization method for efficient design exploration of linear structures under random loads
Computer Methods in Applied Mechanics and Engineering ( IF 6.9 ) Pub Date : 2024-08-27 , DOI: 10.1016/j.cma.2024.117312 Lili Weng , Cristóbal H. Acevedo , Jiashu Yang , Marcos A. Valdebenito , Matthias G.R. Faes , Jianbing Chen
Computer Methods in Applied Mechanics and Engineering ( IF 6.9 ) Pub Date : 2024-08-27 , DOI: 10.1016/j.cma.2024.117312 Lili Weng , Cristóbal H. Acevedo , Jiashu Yang , Marcos A. Valdebenito , Matthias G.R. Faes , Jianbing Chen
Reliability-based design optimization (RBDO) provides a promising approach for achieving effective structural designs while explicitly accounting for the effects of uncertainty. However, the computational demands associated with RBDO, often due to its nested loop nature, pose significant challenges, thereby impeding the application of RBDO for decision-making in real-world structural design. To alleviate this issue, an approximate decoupled approach is introduced for a class of RBDO problems involving linear truss structures subjected to random excitations, with the failure event defined by compliance. This contribution aims to provide an approximate but efficient way for design exploration to facilitate decision-making during the initial design phase. Specifically, the proposed approach converts the original RBDO problem into a deterministic optimization problem through a modest number of reliability analyses by the probability density evolution method (PDEM). Once the deterministic optimization problem is obtained, the solution of the whole RBDO problem can be obtained by solving this equivalent problem without further reliability analysis, resulting in notable enhancement in terms of computational efficiency. In this way, this contribution expands the frontier of application of the operator norm theory within the RBDO framework. Numerical examples are conducted to illustrate the effectiveness and capabilities of the proposed approach.
中文翻译:
一种基于近似解耦可靠性的设计优化方法,用于随机载荷下线性结构的高效设计探索
基于可靠性的设计优化 (RBDO) 提供了一种很有前途的方法,可以实现有效的结构设计,同时明确考虑不确定性的影响。然而,与 RBDO 相关的计算需求,通常是由于其嵌套循环的性质,带来了重大挑战,从而阻碍了 RBDO 在实际结构设计中应用于决策。为了缓解这个问题,为一类涉及线性桁架结构的 RBDO 问题引入了一种近似解耦方法,其中失效事件由柔度定义。此贡献旨在为设计探索提供一种近似但有效的方法,以促进初始设计阶段的决策。具体来说,所提出的方法通过概率密度进化法 (PDEM) 进行适度数量的可靠性分析,将原始的 RBDO 问题转化为确定性优化问题。一旦得到确定性优化问题,就可以通过解决这个等效问题来获得整个 RBDO 问题的解,而无需进一步的可靠性分析,从而在计算效率方面得到显着提高。通过这种方式,这一贡献扩展了 RBDO 框架内算子范数理论的应用前沿。通过数值示例来说明所提出的方法的有效性和能力。
更新日期:2024-08-27
中文翻译:
一种基于近似解耦可靠性的设计优化方法,用于随机载荷下线性结构的高效设计探索
基于可靠性的设计优化 (RBDO) 提供了一种很有前途的方法,可以实现有效的结构设计,同时明确考虑不确定性的影响。然而,与 RBDO 相关的计算需求,通常是由于其嵌套循环的性质,带来了重大挑战,从而阻碍了 RBDO 在实际结构设计中应用于决策。为了缓解这个问题,为一类涉及线性桁架结构的 RBDO 问题引入了一种近似解耦方法,其中失效事件由柔度定义。此贡献旨在为设计探索提供一种近似但有效的方法,以促进初始设计阶段的决策。具体来说,所提出的方法通过概率密度进化法 (PDEM) 进行适度数量的可靠性分析,将原始的 RBDO 问题转化为确定性优化问题。一旦得到确定性优化问题,就可以通过解决这个等效问题来获得整个 RBDO 问题的解,而无需进一步的可靠性分析,从而在计算效率方面得到显着提高。通过这种方式,这一贡献扩展了 RBDO 框架内算子范数理论的应用前沿。通过数值示例来说明所提出的方法的有效性和能力。