基于协同优化的耐压柱壳多学科多目标优化
Multidisciplinary and multiobjective optimization of pressurized cylindrical shell based on collaborative optimization
王萌 1刘峰 1田震 1杨松源1
作者信息
- 1. 哈尔滨工程大学 船舶工程学院,黑龙江 哈尔滨 150001
- 折叠
摘要
为提高耐压柱壳的设计效率和结构性能,本文基于多学科设计优化思想,进行了耐压柱壳多学科设计优化的学科分解,确定了耐压柱壳多学科优化设计的设计参数,明确了设计变量间的耦合关系,设计了耐压柱壳参数化分析流程,进行了设计变量灵敏度分析.在利用响应面模型建立了具有较高拟合精度耐压柱壳近似模型的基础上,以质量排水量比和极限载荷为目标函数,建立了基于协同优化方法的耐压柱壳多学科多目标优化模型,在确定了系统层和学科层的优化求解方法基础上,设计了基于协同优化的耐压柱壳多学科多目标优化框架,进行了优化的求解工作,得到了优化方案.与初始方案的对比表明:所得到的Pareto 解集中,部分方案在 2 个目标函数上均得到了优化,验证了优化框架的有效性.
Abstract
To improve the design efficiency and structural performance of the pressurized cylindrical shell,the mul-tidisciplinary design optimization(MDO)of the pressurized cylindrical shell was decomposed to determine the MDO design parameters and the coupling relationship between the design variables.The parametric analysis of the pres-surized cylindrical shell was designed,and the sensitivity of the design variables was analyzed.Based on the ap-proximation model of a pressurized cylindrical shell with high fitting precision established using a response surface model,with the weight displacement ratio and ultimate load as the objective functions,a multidisciplinary and mul-tiobjective optimization model for the pressurized cylindrical shell based on collaborative optimization was estab-lished.Therefore,the optimized solution method of the system layer and the discipline layer was determined,and a multidisciplinary and multiobjective optimization framework of the pressurized cylindrical shell based on collabora-tive optimization was further designed to optimize the solution,thereby obtaining the optimization scheme.The com-parison with the initial scheme showed that some schemes in the Pareto solution set were optimized on two objective functions,which verifies the effectiveness of the optimization framework.
关键词
耐压柱壳/设计参数/耦合关系/参数化分析流程/近似模型/目标函数/协同优化方法/多学科多目标优化/优化框架Key words
pressurized cylindrical shell/design parameter/coupling relationship/parametric analysis process/approximation model/objective functions/collaborative optimization method/multidisciplinary and multiobjective optimization/optimization framework引用本文复制引用
基金项目
国家自然科学基金(51709063)
中央高校基本科研业务费专项(HEUCFP201603)
中央高校基本科研业务费专项(HEUCFP201735)
出版年
2024
国家科技期刊平台
NSTL
万方数据