Abstract: In order to solve the problem of incomplete preference information in large-group emergency decision-making, a new method of incomplete risk preference information for large-group emergency decision-making is proposed in this paper. Firstly, the optimal discrete fitting method is used to measure the risk preference factors of decision makers and then cluster them. Secondly, according to incomplete preference matrices, attribute correlation measure is carried out, and a new complementary value model based on risk preference and attribute association is proposed to obtain the complete preference information matrices. Thirdly, the principal component of attributes is extracted by principal component analysis method, aggregating information and selecting alternatives are carried out by combining attributes' weight. Finally, the feasibility and effectiveness of the proposed method are verified by the super emergency of typhoon sky pigeon.

Key words: attribute association, risk preference, incomplete preference information, large-group, emergency decision-making

摘要： 针对特大突发事件应急决策中大群体专家存在偏好信息不完全的问题,本文提出一种新的不完全风险性信息大群体应急决策方法。首先,利用最优离散拟合方法对决策者的风险偏好因子进行测度并据此对专家聚类;其次,根据不完全偏好矩阵进行属性关联测度,提出了基于风险偏好和属性关联的新的补值模型,得到完全偏好信息矩阵;然后,运用主成分分析方法提取属性主成分,并结合属性权重进行信息集结和方案择优;最后,通过台风“天鸽”事件验证所提方法的可行性和有效性。

关键词: 属性关联, 风险偏好, 不完全信息, 大群体, 应急决策