三相电流主分量融合的电机故障图形诊断方法

Graphic diagnosis method of motor faults based on components fusion of three phase currents PCA

为了在线快捷有效检测识别电机运行状态,并准确诊断交流电机常见故障,将图形融合方法与主分量分析方法(PCA)相结合,充分利用电机三相定子电流信号实现了电机故障的图形化在线检测方法。应用主分量分析法将三维电流状态空间样本进行降维并在二维平面内实现主分量的图形化融合,定义了电机正常、定子匝间短路、转子断条等不同运行状态下的状态识别实用判据,并针对主分量融合图形定义了故障严重度指标。通过理论分析及电机故障试验,论证了所提诊断方法对电机常见定子匝间短路及转子断条故障检测及诊断的直观性、准确性及载荷变化的鲁棒性,为电机及其拖动系统的故障诊断提供新的思路和方法。

In order for on-line efficient detection and accurate diagnosis of common faults of AC motors, an approach is presented. Dimension of current data sample matrix were reduced by using principal component analysis（ PCA） algorithm and formed 2-D figure. Practical,useful criteria of state identification was defined for the normal condition, stator winding inter-turn short circuit fault, rotor broken bar fault of motor. At the same time, fault severity index of motor was defined too. The characteristic of intuitive, accuracy and robustness of the fault diagnosis algorithm was verified by way of theoretical analysis and fault motor experiments. The method provides a new thought and method for fault diagnosis of motors and driving systems.