中文摘要：

针对旋转机械的故障属性均为模糊描述,提出基于模糊多属性决策的故障诊断专家系统推理机模型.其推理过程为：由频率分析得到故障假设集,通过人机接口获取属性,建立决策矩阵,熵法权值分配,折衷型方法,计算各故障假设集与故障样本间属性的模糊距离,对故障假设集重新排序,完成故障的自动诊断.

英文摘要：

In order to solute the description of the {allure attributes of rotating machinery is fuzzy completely; this paper developed the inferential modal which was based on the fault diagnosis expert system and the Fuzzy Multiple Attribute Decision Making （FMADM）. In this model the decision making matrix is built, on the basis of determining the assumptive fault matrix through frequency analysis and getting the description of attributes from man-machine interface. The additive weighting and the compromise model were used to compute the fuzzy distances between every assumptive {ault matrix and the fault sample matrix. According to these distances value, the assumptive fault matrix is reconstructed to give the result of fault diagnosis automatically.