上海交通大学学报(自然版) ›› 2014, Vol. 48 ›› Issue (04): 493-497.

• 交通运输 • 上一篇    下一篇

基于径向基神经网络模型的耐压壳6σ设计

程妍雪1,庞永杰1,杨卓懿1,冯亮2
  

  1. (1.哈尔滨工程大学 船舶工程学院,哈尔滨 150001; 2.中国海洋大学 工程学院,山东 青岛 266100)
     
     
     
  • 收稿日期:2013-09-18
  • 基金资助:

    国家自然科学基金资助项目(51009040\\E091002),国家高技术研究发展计划(863)项目(2011AA09A106)

6σ Design for Pressurized Cylindrical Shells Based on RBF
 

CHENG Yanxue1,PANG Yongjie1,YANG Zhuoyi1,FENG Liang2
  

  1. (1. College of Shipbuilding Engineering, Harbin Engineering University, Harbin 150001, China;2. College of Engineering, Ocean University of China, Qingdao 266100, Shandong, China)
  • Received:2013-09-18

摘要:

为了得到更可靠的潜水器耐压壳结构,考虑到材料性能和加工时结构尺寸误差的不确定性对结构性能的影响,将6σ设计的思想引入耐压壳优化设计.借助多学科优化平台iSIGHT,对耐压壳进行6σ分析的优化设计,得到优化且可靠的设计方案.为减少6σ优化设计中的计算代价,采用足够精确的径向基神经网络模型代替ANSYS仿真程序进行迭代.结果表明,在保证计算结果足够精确的前提下较大节省了计算时间.
 
 

关键词: 径向基神经网络模型, 耐压壳优化, 6&sigma, 设计, 可靠性

Abstract:

To obtain a reliable pressure hull structure, the 6σ design was introduced into the optimization of pressure hull. The influences of the uncertainty of material and dimension error when processing were fully taken into consideration. Based on the MDO platform iSIGHT, a reliable optimal plan involving 6σ was obtained. In order to reduce the calculation expense, a sufficiently accurate RBF model was used to replace the simulation program ANSYS when iterating, which not only ensured the accuracy but also saves the calculating cost.
Key words:

Key words: pressure hull optimization; 6&sigma, design; reliability; radical basis function(RBF) model

中图分类号: