基于关联维数改进算法滚动轴承故障诊断  

Fault diagnosis of the rolling bearing based on correlation dimension improved algorithm

作　　者：康静秋[1] 孙芳[1] 柳亦兵[1] 

机构地区：[1]华北电力大学自动化系,北京102206

出　　处：《仪器仪表学报》2006年第z1期404-406,共3页Chinese Journal of Scientific Instrument

摘　　要：将分形理论应用于机械故障诊断领域,并针对实际应用中存在已久的关联维数计算不准和不好测评的问题,提出自动计算关联维数的改进算法。该算法在嵌入维数和延迟时间的参数选择上做了改进,并实现自动判别线性标度区。文中将改进的关联维数算法用于滚动轴承振动的数据,验证了关联维数自动算法作为故障诊断特征提取方法的可行性。Fractal theory is applied to the machine fault diagnosis.An improued algorithm is introduced for the correlation dimension estimation in an automatic way,which can avoid the imprecise calculation of the correlation dimension.This new algorithm implements automatically distinguishing of the scaling region.As a typical example this algorithm was applied to rolling bearing condition monitoring.Analysis result demonstrates that the correlation dimension is able to identify a bear fault conditions.

关 键 词：分形 故障诊断 关联维数 滚动轴承 

分 类 号：TH7-55[机械工程—仪器科学与技术]