主管：中国科学院
主办：中国优选法统筹法与经济数学研究会
　　　中国科学院科技战略咨询研究院
中国管理科学

中国管理科学 ›› 2004, Vol. ›› Issue (1): 64-67.

• 论文 • 上一篇    下一篇

动态多指标决策问题的投影寻踪模型

金菊良1, 汪淑娟1, 魏一鸣2   

  1. 1. 合肥工业大学土木建筑工程学院, 安徽, 合肥, 230009;
    2. 中国科学院科技政策与管理科学研究所, 北京, 100080
  • 收稿日期:2003-03-05 修回日期:2003-11-20 出版日期:2004-02-28 发布日期:2012-03-07
  • 基金资助:
    教育部优秀青年教师资助计划(教人司[2002]350);四川大学高速水力学国家重点实验室开放基金(0201);安徽省优秀青年科技基金;安徽省自然科学基金资助项目(01045102)

Projection Pursuit Model for Dynamic Multiple Attribute Decision Problems

JIN Ju-liang1, WANG Shu-juan1, WEI Yi-ming 2   

  1. 1. College of Civil Engineering, Hefei University of Technology, Hefei 230009, China;
    2. Institute of Policy & Management, Chinese Academy of Sciences, Beijing 100080, China
  • Received:2003-03-05 Revised:2003-11-20 Online:2004-02-28 Published:2012-03-07

PDF (PC)

2186

被引次数

104

摘要： 针对动态多指标决策中指标和时段的权重确定问题,提出了基于投影寻踪(PP)的理想点法新模型(PP-IPM模型)。该模型利用决策矩阵样本的内部信息,把方案的三维决策矩阵综合成一维投影值,投影值越大表示该方案越优,根据投影值的大小就可对各方案进行综合排序决策。建议用实码加速遗传算法进行PP-IPM的建模,简化了PP技术的实现过程,克服了目前PP技术计算过程复杂、编程实现困难的缺点。实例计算的结果说明,直接由决策矩阵样本数据驱动的PP-IPM模型用于动态多指标决策问题简便可行,适用性和可操作性强,具有推广应用价值。

关键词: 动态多指标决策, 决策矩阵, 排序, 理想点法, 投影寻踪, 遗传算法

Abstract: In order to determine the weights of attributes and period times in dynamic multiple attribute decision,an ideal point method based on projection pursuit for dynamic multiple attribute decision,named PP-IPM,is presented in this paper.In PP-IPM model,by using the interior information of decision matrix,decision matrices of decision schemes in three dimensions can be synthesized with projection values in one dimension which indicates comprehensive quality of decision schemes,and the decision schemes can be ordered according to the projection value of each scheme.By using real coding based accelerating genetic algorithm developed by the authors,the modeling of PP-IPM can be simplified by the realized process of projection pursuit technique,and can overcome the shortcomings of large computation amount and difficulty of computer programming in traditional projection pursuit methods.The case study shows that applying PP-IPM model driven directly by decision matrix samples data is simple,feasible,general and practical,and PP-IPM model can be applied to many dynamic multiple attribute decision problems.

Key words: dynamic multiple attribute decision, decision matrix, sort order, ideal point method, projection pursuit, genetic algorithm

中图分类号: