基于EMD的水轮机故障识别

The Research on Faults Recognitions of Turbine Based on EMD

利用EMD方法可依据信号本身固有特性进行分解的特征,对水力发电机组振动测试信号进行分析,并与傅里叶变换等方法进行比较。研究表明:基于EMD方法可以得到机组振动信号的IMF分量,能量占据主导地位的IMF分量与机组运行状态密切相关,该IMF分量直接由故障源诱发,因而采用EMD方法能够准确地捕捉到机组振动信号中所蕴含的真实物理意义,准确识别机组运行状态。

According to the intrinsic mode of testing signals, the vibration characteristics of signals is analyzed based on EMD, and the results is compared with that of other methods. And the research show that the IMF of signals is result of multi - band EMD filtering, and the dominate IMF is closely related to the running state of hydraulic unit, so using EMD method can capture the physical meaning of vibration signals to identify the running type of unit.