基于最优属性值的混合多属性动态VIKOR决策方法

Extended VIKOR method for dynamic hybrid multi-attribute decision making based on optimum attributes values

针对传统VIKOR方法不能直接处理多类型属性的问题,提出了基于最优属性值的混合多属性的动态决策方法,给出了各类型属性值之间距离的定义;根据各方案属性值与最优属性值的相对距离来计算各时段的"群体效益"值和"个别遗憾"值,该方法可直接用于多种类型的属性和属性值;通过逼近最优属性值的方法获取属性权重,采用权重离差之和最小的时间赋权方法,集结不同时段的评价值,对计算过程中遇到的特殊情况进行了分析。实例验证表明:该方法简便有效且更能体现决策者的主观偏好。

The traditional VIKOR method cannot deal with the problem of multiform attributes, so a dynamic evaluation method is proposed to handle the multiform attributes and multiform attribute values based on the optimum attribute values. The distance between the multiform attributes is de- fined. The group utility measurement and individual regretting measurement are made according to the relative distance between the alternatives attribute values and the optimum attribute values. This method can be directly used to deal with multiform attributes and multiform attribute values. The weights of attributes are computed in a way to approach the optimum attribute values. The evaluation values of different times are integrated by the time-weighted method that minimizes the sum of devia- tion among time weights, and are used to analyze the particular case appearing in the process. The real example indicates that the proposed method is more convenient and effective and can embody the preference of a decision maker well.