基于决策者偏好构造的概率语言型Z-number多属性决策方法

由于决策环境的日益复杂，专家评价信息往往会包含专家偏好，因此有必要对专家偏好进行识别与处理。针对这一情况，提出了基于决策者偏好构造的概率语言型Z-number并应用于多属性决策问题中。首先，针对专家偏好未知的情况，提出了一种基于修正的平均得分距离，由专家评价信息确定专家偏好。然后，依据得到的专家偏好，利用GOWPA算子与最大熵生成概率语言型Z-number。同时，获得专家评价信息的可信度，由专家可信度聚合专家信息并采用TODIM方法对方案进行排序。最后，在国家大力提倡节能减排的背景下，将所提方法应用于国家绿色矿山试点企业的选取上，验证了所提方法的有效性和可行性。

A Probabilistic Linguistic Z-number Multi-attribute Decision Making Method Based on Decision Maker Preference

Due to the increasingly complex decision-making environment, expert evaluation information usually contains expert preferences, it is necessary to identify and process expert preferences. To solve this problem, a probabilistic linguistic Z-number based on decision maker''s preference is proposed and applied to multi-attribute decision making problems. Firstly, for the situation of unknown expert preference, a modified mean score distance based on expert evaluation information is proposed to determine expert preference. Then, according to the obtained expert preference, the GOWPA operator and maximum entropy are used to generate the probabilistic linguistic Z-number. At the same time, the credibility of the expert evaluation information is obtained, the expert information is aggregated by the expert credibility, and the schemes are sorted by the TODIM method. Finally, under the background of vigorously advocating energy conservation and emission reduction, the proposed method is applied to the selection of national green mine pilot enterprises to verify the effectiveness and feasibility of the proposed method.