基于后悔理论和DEMATEL的语言型多属性决策方法

针对目前语言型多属性决策方法大多基于期望效用理论且不考虑指标间影响关系的不足,提出了一种将后悔理论和决策试验与评价试验法相结合的语言型多属性决策方法。首先,依据后悔理论的思想,定义了语言后悔-欣喜函数,给出了方案感知效用值的计算公式;然后,利用决策试验与评价试验法分析指标间的影响关系,给出了基于语言DEMATEL的指标权重确定方法,再通过指标总容量最大优化模型给出了基于注水原理的指标权重确定方法,并在此基础上求解方案的综合感知效用值,据此对方案进行排序择优;最后,通过两个应用实例来验证所提方法的可行性和有效性。实例结果表明,由于该方法同时考虑了决策者的心理行为和指标间的影响关系,因此可使决策结果更加贴近现实且更为可靠。

A New Linguistic Multiple Attribute Decision Making Method Based on Regret Theory and DEMATEL

In the face of some complex multiple-attribute decision-making problems, the decision maker prefers to give linguistic evaluation information such as good, medium, bad, et al, and therefore linguistic multiple-attribute decision-making problem occurs. Linguistic multiple-attribute decision-making problem can be seen everywhere in real life, and has attracted considerable attention from domestic and overseas experts and scholars in recent years. However, so far the linguistic multiple-attribute decision-making methods are mostly based on expected utility theory, and do not consider the influence relationships among indexes. To overcome these defects, a new linguistic multiple-attribute decision-making method that combines the regret theory, Decision Making Trial and Evaluation Laboratory (DEMATEL) is proposed, in which the decision maker’s regret aversion behavior and the influence relationships among indexes are considered. Firstly, a linguistic regret-rejoice function is defined on the base of the regret theory and the computational formula of perceived utility value is given. Secondly, the index weights based on linguistic DEMATEL are calculated through analyzing the influence relationships among indexes, and the index weights based on water-filling theory are calculated through a programming model that is constructed by the maximum total capacity of index, then the comprehensive perceived utility value is obtained, through which we can rank the alternatives. Finally, two examples are used to illustrate the feasibility and validity of the proposed method. The results show that this method is more practical and reliable as it considers the decision maker’s psychological behavior and the influence relationships among indexes. This paper expands the study of the traditional linguistic multiple-attribute decision-making method, and provides an important reference and guidance for solving linguistic multiple-attribute decision-making problem.