基于迭代均值组合近似模型和序贯优化与可靠性评估法的船舶结构优化设计

发布时间：2018-05-25 02:05

本文选题：船舶结构 + 迭代均值组合近似模型　；参考：《上海交通大学学报》2017年02期

【摘要】：针对传统可靠性优化方法不能同时保证大型船舶结构计算效率与精度的问题,在船舶结构优化设计中考虑设计变量的不确定性,通过迭代均值组合(ER)近似模型和粒子群算法(PSO)提高序贯优化与可靠性评估(SORA)法的计算效率和精度,建立了改进的可靠性优化设计流程并解决了多用途船舱段的可靠性优化问题.其中:SORA法将可靠性分析与优化过程解耦,并提高了可靠性优化效率;PSO保证了SORA法优化所获全局最优解的计算效率;ER近似模型能够以最少的采样点建立满足精度要求的近似模型,并代替最有可能的失效点(MPP)和优化过程中耗时的有限元计算.多用途船舱段的可靠性优化结果表明:可靠性优化相较于确定性优化更有实际意义;改进的可靠性优化设计流程在保证计算精度的同时能够显著降低计算成本.
[Abstract]:In order to solve the problem that the traditional reliability optimization method can not simultaneously guarantee the efficiency and accuracy of the large ship structure calculation, the uncertainty of the design variables is considered in the optimization design of the ship structure, and the efficiency and accuracy of the sequential optimization and reliability evaluation (SORA) method are improved by the iterative mean combination (ER) approximation model and the particle swarm optimization (PSO) algorithm. The improved reliability optimization design process is established and the reliability optimization problem of multi purpose cabin is solved. The SORA method decoupling the reliability analysis and optimization process and improving the efficiency of reliability optimization; PSO ensures the calculation efficiency of the global optimal solution obtained by the SORA method, and the ER approximation model can be established with the least sampling point. The approximate model that meets the requirements of the precision is replaced by the most likely failure point (MPP) and the time-consuming finite element calculation in the optimization process. The reliability optimization results of the multi purpose cabin section show that the reliability optimization is more practical than deterministic optimization, and the improved reliability optimization design process can show the calculation accuracy while ensuring the accuracy. Reduce the cost of calculation.
【作者单位】：上海交通大学海洋工程国家重点实验室;高新船舶与深海开发装备协同创新中心;
【基金】：教育部重大专项基金项目(201335)资助
【分类号】：U662;TP18

【相似文献】