A reliability-based multidisciplinary design parallel optimization method based on double-layer approximation model for nuclear fuel assembly bottom nozzle

Changfu Wan,Wenqiang Li,Bo Yang,Sitong Ling,Yingdong Liu
DOI: https://doi.org/10.1016/j.pnucene.2024.105292
IF: 2.461
2024-05-30
Progress in Nuclear Energy
Abstract:The design of intricate systems such as the nuclear fuel assembly necessitates elevated levels of multi-disciplinary performance and reliability. In current reliability-based multidisciplinary design optimization methods, the decoupling strategy between optimization and reliability assessment involves sequential optimization and reliability assessment, with reliability serving as a constraint for optimization. This approach often suffers failure in locating the global optimal solution. In light of this issue, a reliability-based multidisciplinary design parallel optimization method based on double-layer approximation model is hereby proposed, achieving parallel separation and global efficient optimization of deterministic disciplinary performances and reliability, while performing efficient and accurate calculation on the multi-disciplinary reliability numerical solution. This method has been implemented for the multi-disciplinary uncertain optimization of the nuclear fuel assembly bottom nozzle. In comparison to conventional optimization methods, this approach demonstrates enhanced global optimization effectiveness.
nuclear science & technology
What problem does this paper attempt to address?