A hybrid approach for reliability-based robust design optimization of structural systems with dependent failure modes

Hao Lu,Zhencai Zhu,Yimin Zhang
DOI: https://doi.org/10.1080/0305215X.2019.1586893
IF: 2.5
2019-10-07
Engineering Optimization
Abstract:Estimating system reliability has been recognized as a critical step in reliability-based design optimization for structural systems. The estimation of system reliability entails multi-dimensional integration, which is very difficult and numerically expensive. The proposed hybrid method integrates the concepts of saddlepoint approximation, the dimension reduction method (DRM) and copulas. The moment-based saddlepoint approximation enhanced by the DRM is used to calculate the component reliability. The copula method is adopted to realize the joint failure probability modelling between two failure modes. Reliability-based sensitivity analysis for the reliability-based robust design optimization is carried out as well. The structural system optimization model is established, including the Euclid norm of the reliability sensitivity as an optimization objective and system reliability as a constraint. Examples demonstrate that the optimization results obtained by the proposed approach ensure the maximization of the robustness of the structural system.
What problem does this paper attempt to address?