东北大学学报（自然科学版）

东北大学学报:自然科学版 ›› 2015, Vol. 36 ›› Issue (6): 908-912.DOI: 10.12068/j.issn.1005-3026.2015.06.032

• 管理科学 • 上一篇    

基于案例的复杂产品快速优化设计

冯国奇1， 崔东亮2， 马明旭3   

  1. (1.东北大学 工商管理学院， 辽宁 沈阳110819； 2.东北大学 流程工业综合自动化国家重点实验室， 辽宁 沈阳110819; 3.东北大学 机械工程与自动化学院， 辽宁 沈阳110819)
  • 收稿日期:2014-04-30 修回日期:2014-04-30 出版日期:2015-06-15 发布日期:2015-06-11
  • 通讯作者: 冯国奇
  • 作者简介:冯国奇(1976-)，女，河南平顶山人，东北大学副教授.
  • 基金资助:
    国家自然科学基金资助项目(71102120)； 中央高校基本科研业务费专项资金资助项目(N130406002， N130408001).

Case-based Fast Optimization Design for Complex Products

FENG Guo-qi1， CUI Dong-liang2， MA Ming-xu3   

  1. 1. School of Business Administration， Northeastern University， Shenyang 110819， China; 2. State Key Laboratory of Synthetical Automation for Process Industries， Northeastern University， Shenyang 110819， China; 3. School of Mechanical Engineering & Automation， Northeastern University， Shenyang 110819， China.
  • Received:2014-04-30 Revised:2014-04-30 Online:2015-06-15 Published:2015-06-11
  • Contact: FENG Guo-qi
  • About author:-
  • Supported by:
    -

RichHTML

0

PDF (PC)

707

摘要： 充分挖掘设计案例中蕴含的规律以提升产品研发效率和质量，是复杂产品设计企业持续研究的关键问题之一.跳出多学科设计优化的传统思维方式，从知识应用角度出发提出一个产品快速优化设计方案:以源于设计方案的关键参数矢量为基础，通过求解输入/输出参数之间的关系完成权值设计，基于专家经验及设计目标实现快速聚类分析，通过类内参数延拓和抽样完成参数的高覆盖、合理化设计，采用试验设计的方法为计算复杂度约束的优化设计提供高质量初始解，对全排列的其他样本进行基于知识的评估以快速筛选较优方案进行仿真.实验结果表明了本方法的有效及可操作性.

关键词: 复杂产品, 设计知识, 优化设计, 加权聚类

Abstract: Optimization design for complex products is of great difficulty because of its multidisciplinary characteristics. Multidisciplinary design optimization (MDO) can better deal with this issue， but there are still some flaws in implementing this technology. The exploration of the laws from previous cases can enhance the efficiency and quality of product design. Unlike the traditional methods of MDO， a fast design optimization is proposed from the perspective of knowledge application. Based on the key parameters extracted from design versions， the weights of vector are acquired by the calculation of the relationship between input parameters and output parameters. Rapid clustering is performed based on expert experience and design objectives， parameters are designed by the extension and re-sampling of cluster members， experiments are designed to provide high-quality initial solutions to the optimization designs with computational constraint， and the other samples from the same cluster are given knowledge-based evaluation and simulated with fast sampling optimization scheme. The results showed that this method is of great effectiveness and feasibility.

Key words: complex product, design knowledge, optimization design, weighted clustering

中图分类号: