Abstract: In view of two problems existing in FAHP, one is rank reversal problem and another is the limited capability in dealing with uncertainty. When it is applied to evaluation and selection for product schemes, this paper intends to treat criterion weights and performance scores of alternatives on criteria separately and introduce Evidential Reasoning（ER） theory, which is good at uncertain reasoning, into the systematic framework of FAHP to calculate the performance scores of alternatives. Then a hybrid group based decision model FAHP-ER is proposed by combining both FAHP and ER approach together, which enhances capability of FAHP in dealing with uncertainty, another weakness in FAHP, due to introduction of ER algorithm. A case study for rotor and bearing system scheme evaluation is presented to demonstrate the application of hybrid decision model and finally a reasonable rotor system scheme is obtained.

Key words: fuzzy Analytic Hierarchy Process（AHP）, rank reversal, evidential reasoning theory, scheme evaluation, rotor and bearing system

摘要： 针对模糊层次分析法（Fuzzy AHP，FAHP）用于产品方案评价时存在的逆向排序问题和不确定信息处理问题，分析其原因后将不确定性推理的证据推理（Evidential Reasoning，ER）理论引入FAHP的层次结构进行底层方案评价值的计算，在此基础上提出了FAHP-ER混合决策模型，该模型由于ER的引入而大大提高了它在不确定信息处理方面的能力，从而克服了FAHP方法对不确定信息处理不足的问题。用一个轴承转子系统设计方案的评价实例对混合决策模型进行了验证，很好地处理了方案评价过程中决策者的各种主观不确定信息，在此基础上获得了最佳的转子设计方案。

关键词: 模糊层次分析法, 逆向排序, 证据理论, 方案评价, 轴承转子系统