基于粗糙集理论的振动信号数据挖掘研究  

Research on Data Mining of Vibratory Signal Based on Rough Set Theory

作　　者：薛庆增[1] 李冬[2] 孙涛[3] 邓斌[2] 孙波[2] 

机构地区：[1]海军驻沈阳导弹专业军事代表室,辽宁沈阳110043 [2]海军航空工程学院研究生管理大队,山东烟台264001 [3]海军航空工程学院飞行器工程系,山东烟台264001

出　　处：《航空计算技术》2010年第6期50-54,共5页Aeronautical Computing Technique

摘　　要：利用转子试验台,采集了转子不对中、支座松动和转子碰磨等三种振动信号。以采集到的部分振动信号为例,在时域上进行小波变换的能量特征提取。并以能量为特征参数,利用粗糙集理论对采集到的信号进行数据挖掘和获取故障规则。在获取的规则基础上,对余下的振动信号进行了验证,结果表明诊断全部正确,规则是有意义的。This paper collectes the vibratory signals of rotor misalignment fault,pedestal looseness fault and rub between rotor and stator fault by rotor experimental platform.As the example of parts of the vibratory signal collected,the energy feature is extracted by Wavelet Transform in the field of time.The data mining and fault rule of signals collected are gained by rough set theory as the feature parameter of energy.The remaining vibratory signal is validated on the basis of rule,the results indicate the diagnosis is correct completely,and the rule is meaningful.

关 键 词：振动信号 粗糙集 数据挖掘 

分 类 号：O242[理学—计算数学]