基于电机定子电流的齿轮故障诊断方法
Gear fault diagnosis method based on MCSA

摘要 齿轮故障诊断一般采用振动信号进行故障特征提取,但振动诊断法不便于安装传感器,易受环境和噪声影响．电机本身具有传感器的特性,定子相电流等信号能够反映负载转矩的变化．因此,针对由电机驱动的齿轮传动系统,提出了一种基于电机定子相电流分析的齿轮无损故障诊断方法．推导了电机定子电流如何反应负载转矩的特性,并分析了齿轮正常与故障状态下定子电流的频谱特征,发现可通过观察边频带的出现来判断齿轮发生局部式故障．通过Matlab对故障诊断原理进行了仿真验证,在实验平台上结合频谱分析成功检测出了齿轮断齿故障．
Abstract： Traditionally, the vibration signal is used for gear fault diagnosis. However, the mechanical sensors are not convenient to be installed, and may be influenced by the environments and noise. Some electrical signals such as the motor current signal can track the pulsations of load torque. So the motor can be considered as a torque sensor. This paper proposes a gear fault diagnosis method based on motor current signature analysis (MCSA). Firstly, the mechanism of how stator current reflects the characteristics of load torque is deduced. Also, the motor current spectrums under the gear without fault and with one broken tooth are analysed. When the gear has a local fault, a series of frequency band can be observed. Then this phenomenon is verified in Matlab. The gear fault is finally successfully detected on the experimental platform combined with spectrum analysis.