DOC《多维随机不确定性下的船舶多学科》

多维随机不确定性下的船舶多学科 稳健设计优化研究 李冬琴1,2,蒋志勇1,2,赵欣1 (1.

江苏科技大学 船舶与海洋工程学院,江苏镇江 212003;

2. 江苏现代造船技术有限公司,江苏镇江 212003) 摘要:通常船舶多学科设计优化仅考虑确定性的影响,忽视设计中不确定性因素的存在,显然优化结果无法真正保障船型方案的可行性和航行性能的优良性.目前,不确定性船舶多学科设计优化主要集中在一维不确定性分析.为了得到船舶设计优化的最优方案,以多维不确定性下的多学科稳健设计优化问题为例,通过分析考虑随机不确定性因素的均匀分布和正态分布并存分布状况,提出了一种新的以Legendre polynomials和Hermite polynomials多项式为基础的多维多项式混沌方法,并以海洋平台支持船为对象构建多维不确定性量化模型.通过对多维随机不确定性因素对船舶优化方案的影响分析完成了船舶多学科稳健设计优化研究,有效地减少和避免船舶设计优化方案失效的可能性. 关键词:随机不确定性;

多学科设计优化;

多维多项式混沌方法;

不确定性量化 中图分类号: 文献标志码:A 【DOI】10.13788/j.cnki.cbgc.2015.11.001 Ship Multidisciplinary Robust Design Optimization under the Multidimensional Stochastic Uncertainties LI Dong-qin1,2, JIANG Zhi-yong 1,2, Zhao Xin1 (1.School of Naval Architecture &

Ocean Engineering, Jiangsu University of Science and Technology, Jiangsu Zhenjiang 212003, China;

2.Jiangsu Modern Shipbuilding Technology, Ltd., Jiangsu Zhenjiang 212003, China) Abstract: Multidisciplinary Design Optimization in ship design only considers the impact of certainties, ignoring the existence of uncertainties in the design. The results of the optimization obviously can'

t really guarantee the feasibility of ship hull and excellence of sailing performance. Currently, Uncertainty Multidisciplinary Design Optimization of ship focuses on one dimension of uncertainty analysis. To obtain the optimal solution of ship design, the problem of Multidisciplinary Robust Design Optimization under multi-dimensional uncertainties was studied in this paper by analyzing and considering uniform distribution and normal distribution of random uncertainties, deduced a new multi-dimensional polynomial chaos approach based on Legendre polynomials and Hermite polynomials,established the mathematical models of multi-dimensional uncertainty quantification while considering offshore supply vessel as the research object. By analyzing the influence of multi-dimensional stochastic uncertainties to the ship design optimization, the research of ship robust multidisciplinary design optimization was completed in this paper;

the possibility of failure in ship design optimization was effectively reduced and avoided. Key words: stochastic uncertainty, Multidisciplinary Design Optimization, multi-dimensional polynomial chaos, uncertainty quantification

0 引言 目前,大多数船舶多学科设计优化模型是建立在确定性多学科优化理论基础之上,实际上船舶最优方案的选择会受到很多不确定性因素的影响,从而导致最优方案可能失效或者船舶性能达不到预定的设计目标.对于考虑不确定性因素存在的多学科设计优化问题,不确定性多学科设计优化(Uncertainty Multidisciplinary Design Optimization,UMDO)方法是有效的解决途径[1].其核心思想是通过综合考虑各个学科间的耦合作用,充分考虑不确定性因素的传递影响,获取系统的整体最优解. 文章以海洋平台支持船的稳健设计优化问题为对象,综合考虑各项船舶性能及实际航行过程中多维不确定性因素的影响,以Legendre polynomials和Hermite polynomials多项式为基础推导了多维多项式混沌方法以解决均匀分布和正态分布并存状况下的不确定性量化分析,保证目标性能针对不确定性因素不敏感从而达到稳健性要求,使得复杂船舶工程系统中采用稳健设计优化进行设计方案的决策更为合理[2].