粗糙集和LSSVM在设备故障诊断中的应用

为了提高旋转设备故障诊断的准确率,提出了基于粗糙集和最小二乘支持向量机(LSSVM)的旋转设备故障诊断方法,讨论了如何进行数据选择、离散及约简方法,用粗糙集提取出旋转设备故障诊断的关键征兆属性,降低数据集的维数将约简属性后的数据集送入最小二乘支持向量机进行故障分类训练。仿真结果表明:采用此方法的故障识别率优于PCA-LSSVM法,分类时间也明显优于LSSVM分类方法。

Application of Roughset and LSSVM in Machinery Fault Diagnosis

The method of fault diagnosis of rotation equipment,which based on roughset and least squares support vector machine is proposed for improving accuracy of fault diagnosis,and in data set,selection,discrete,building and reduction method of decision table is discussed,dimension of data set will be reduced which extracted by rough set property of key symptoms of fault diagnosis of rotating equipment.The data set which is reduced properties is sent into least squares support vector machine for fault classification training,result of fault diagnosis is much better than PCA-LSSVM way,and classification time is also more excellent than classification way of LSSVM.