基于信息融合技术的变频器隐性故障诊断研究

Research on converter recessive fault diagnosis based on information fusion technology

针对变频调速系统中常用的电机诊断方法难以准确判断隐性故障类型的问题,采用基于D-S证据推理的信息融合故障诊断方法,将三相电压传感器和三相电流传感器的判断结果进行融合,以识别变频器不对称运行的故障;将A,B,C三相定子电流信号进行融合,以识别转子断条故障.该方法通过小波包频带能量提取电动机隐性故障特征信息,然后利用D-S证据理论融合计算故障特征信息进行故障识别.诊断测试结果表明,该方法可提高电动机故障诊断的精度,并能满足诊断的实时性要求.

To solve the problem that by using common fault diagnosis method of motor in frequency-conversion speed-regulation system it is difficult to judge fault types,the method of fault diagonisis based on D-S evidence theory is used,which can reduce uncertainty in fault diagnosis.The fault of converter in the asymmetrical condition is identified from analysis result of three-phase voltage sensors and three-phase current sensors.The fault of the induction motors is identified by the fusion result of three phase stator current.Then the D-S evidence theory is used to fuse and calculate fault features data so as to identify fault.The experiment of diagnosis test showed the method improves precision of fault diagnosis of motor and can meet real-time requirement of diagnosis.