基于变精度粗糙集的齿轮故障诊断

Fault diagnosis of gear based on variable precision rough set

提出一种基于变精度粗糙集的齿轮故障诊断模型。首先用自组织特征映射神经网络进行数据的离散化;再由变精度近似依赖度进行属性约简;然后生成故障诊断规则。给出一个齿轮的故障诊断的实例,并与粗糙集的故障诊断模型进行比较。结果表明基于变精度粗糙集方法降低了决策规则的复杂度及规则数量,且提高了故障识别率。

A method of gear fault diagnosis is proposed based on variable precision rough set.The discretizing attribute values are obtained by self-organizing map neural network,Using the property of approximation dependency to reduce the attribues,and then generating the fault diagnosis rules.An example of gear fault diagnosis is given,and comparing with rough set fault diagnosis modle.The result shown that the complexity and number of decision rules are decreased.,and the effeciency of fault recognition is raised.