灰色理论在故障特征量电机状态预测中的应用  

The Motor Fault Prediction Based on the Grey Theory

作　　者：许允之[1] 谭清雄[2] 方磊[1] 褚鑫[1] 贾立敬 陈鲁娜 

机构地区：[1]中国矿业大学信息与电气工程学院,江苏徐州221116 [2]华中科技大学水电学院,湖北武汉430074

出　　处：《实验室研究与探索》2015年第3期47-50,73,共5页Research and Exploration In Laboratory

摘　　要：当电机发生故障时,其电气或机械故障会产生除了基波峰值以外的其他峰值。感应电机的转子断条和气隙偏心等故障,都会影响气隙磁势场,进而影响定子的电流频谱,通过对定子电流频谱的分析即可实现故障诊断。先使用快速傅里叶变换将时域定子电流转变到频域,观察3种情况下的频谱图。发现频谱图在25和75 Hz处的幅值明显不同,再使用改进型的傅里叶变换求出幅值。将这两处幅值的均值作为电机故障情况的特征值,使用灰色理论进行预测,并和实际情况对比。结论表明了结果的准确性,并根据预测结果给出电机维修建议。It is difficult for motor failure prediction to find fault characteristic problems. After in-depth study, the authors found that failure will affect the magnetic field inside the motor, and then be reflected by the spectrum. Good motor current spectrum only has one peak. When the motor fails, its electrical or mechanical failure, other peaks will happen. If an induction motor breaks its rotor bars or an air gap eccentricity faults the air gap magnetic potential field will be affected, so the stator current spectrum will be affected. Therefore, fault diagnosis can be achieved through the stator current spectrum analysis. Using fast Fourier transform, the stator current can he transformed from time domain to frequency domain, three cases of spectrum are observed. It is found that the amplitudes are significantly different at the 25 and 75 Hz, then modified Fourier transform is used to calculate the amplitude. The mean amplitude of these two fault conditions is used as motor characteristic value, then by using the gray theory, the actual situation can be predicted and compared. The conclusions show the good accuracy. And the forecast results can give recommendations for motor repair.

关 键 词：灰色理论 故障特征量 磁场分布 故障诊断 故障预测 

分 类 号：TM307.1[电气工程—电机]