计算机工程

• 人工智能及识别技术 • 上一篇    下一篇

区间数信息下的多属性群决策研究

李 磊,谢小璐   

  1. (江南大学商学院,江苏无锡214122)
  • 收稿日期:2013-10-22 出版日期:2014-10-15 发布日期:2014-10-13
  • 作者简介:李 磊(1959 - ),男,教授、博士生导师,主研方向:人工智能,决策理论与方法;谢小璐,硕士研究生。
  • 基金项目:
    国家自然科学基金资助项目(70371051);浙江省高校人文社科重点研究基地重大招标支撑子项目基金资助项目(RWSKZD04- 2012ZB2);江南大学研究生培养创新工程基金资助项目。

Research on Multi-attribute Group Decision Under Interval Number Information

LI Lei,XIE Xiao-lu   

  1. (School of Business,Jiangnan University,Wuxi 214122,China)
  • Received:2013-10-22 Online:2014-10-15 Published:2014-10-13

摘要: 随着决策系统的不断扩大以及复杂化,评价指标以实数值描述稍显欠妥。研究属性权重为实数值、区间数信息下的多属性群决策问题,建立集结区间数信息的非线性规划模型,采用模拟植物生长算法集结各方案中多位专家给出的区间数偏好,并利用得到的偏好区间数构成群决策偏好矩阵。结合已知权重,引入投影理论得到各方案的综合评价值,从而选出最优决策方案。算例结果表明,模拟植物生长算法不仅计算简便、灵活,并且尊重每个专家的意见,较好地保留了信息的完整性,显示出人工智能算法在信息集结方面的有效性。

关键词: 区间数, 群决策, 植物生长模拟算法, 投影理论, 偏好矩阵, 非线性规划

Abstract: With the continuous expansion and complication of the decision system,evaluation indexes described as the real value are slightly defective. This paper studies multi-attribute group decision that the attribute value information is in the form of the interval numbers and the weight is given by real numbers. It establishes nonlinear programming model which is used to gather the information of interval numbers. Plant Growth Simulation Algorithm(PGSA) is proposed to gather each program on each attribute group decision preference ordering intervals which are given by different experts, while achieving the preference matrix. Projection theory with the weights of the attribute value is introduced to sort and choose the best decision-making program. A numerical example results indicate the PGSA both calculates simply and flexibly and respects the views of each expert in order to keep the integrity of the information. It shows the effectiveness of artificial intelligence algorithms in information gathering.

Key words: interval number, group decision, Plant Growth Simulation Algorithm ( PGSA ), projection theory, preference matrix, nonlinear programming

中图分类号: