摘要
提出一种多属性群决策中同时考虑专家群体对属性主观赋权的偏好和决策者对决策重要性认识的偏好来确定属性权重的方法,能够兼容专家实数型、区间型和语言型等类型的属性权重赋值.首先建立标准属性重要差异矩阵以实现专家对属性赋权的优劣比较和差异规范,定义统一的决策者偏好映射对其进行调整;然后求解各矩阵的排序向量以量化属性的相对重要程度,并与专家权重聚合得到属性权重向量;最后给出了方法的具体步骤,并通过算例说明了该方法的具体应用.
Based on subjective preference of experts and decision-maker,a method of determining attribute weights in multi-attribute group decision-making is presented.This method considers the subjective preference of experts on the attribute weights and of decision-makers on the importance awareness of the decision-making importance,and also can be compatible with main date types such as real number,interval number and language number.The standard attribute importance matrix is established to compare each two experts and quantified their differences.After using a universal decision-maker preference mapping to adjust this matrix,all local priority vectors of matrix are figured out to quantify the alternatives' priority,and then are aggregated with expert weights to an attribute weights vector.The concrete steps of determining attribute weights are specialized.Finally,numeral example shows the feasibility of the developed method.
出处
《控制与决策》
EI
CSCD
北大核心
2010年第11期1645-1650,1656,共7页
Control and Decision
基金
国家自然科学基金项目(70872122)
重庆市科技攻关计划项目(CSTC
2009AC2143)
关键词
属性权重
主观偏好
多属性群决策
Attribute weights
Subjective preference
Multi-attribute group decision-making