考虑多种偏好信息的多阶段方案选择方法

Multi-stage scheme selection method considering multiple preference information

针对方案选择过程中专家倾向于采用不同偏好信息的问题,提出了考虑多种偏好信息的方案选择方法。专家可以从方案集中选择出子集,从效用值、偏好次序、互反判断矩阵和模糊偏好关系四种形式中选择一种对方案子集进行评价;采用转化函数将专家给出的偏好信息转化为偏好向量;根据各个专家的偏好向量集结方案分类结果;将结果用于反馈调整各专家的方案子集;最后通过多个阶段的反馈与调整,改进群体一致性以获得最优方案。将该方法用于某汽车制造商汽车零部件再制造方案的选择,验证了所提方法的有效性。

Aiming at the problem that the experts prefer the different preference information in the process of selecting the scheme, a scheme selection method considering multiple preference information is proposed. The expert can choose a subset from the program set and give the preference information in the form of utility value, preference order, multiplicative preference relations and fuzzy preference relations to evaluate the subset of the program. Transformation functions are used to convert the preference information into the preference vector. According to each expert＇s preference vector to aggregate program classification results to feedback and adjust the program subset of every expert. Finally, the multi-stage feedback and adjustment is adopted to improve the group consistency to obtain the optimal scheme. The method is applied to the selection of a vehicle manufacturer＇s auto parts remanufacturing scheme, which illustrates the effectiveness of the proposed method.