Multiple discriminate analysis and data integration of vibration in rotation machinery

摘要

The Multiple Discriminant Analysis system described provides three parameters: (1) a Dynamic Force Factor (DFF) that characterizes the dynamic forces which act to reduce operational life of the bearing; (2) a Bearing Degradation Factor (BDF) that characterizes the actual condition of the rolling element bearing; and (3) a Life Expectancy Factor (LEF) that characterizes the overall condition of the first two factors. Each factor is configured in scalar form, wherein readings range from acceptable, to caution/degradation, to action required. DFF combines low frequency and high frequency dynamic forcing function discriminants. BDF combines, in this case, four powerful diagnostic bearing fault process discriminants, in a formulaic composition. The composition accurately describes the actual rolling element bearing condition indicating optimum or warning of a potential failure condition. The normally voluminous vibration data is compressed into three easily understood, yet highly informative numbers.