The Fault Diagnosis Of Large Wind Turbines Based On Wavelet Transform

Recently, the energy crisis is getting worse. As a kind of clean renewable resources, wind energy has become the main trend of the energy crisis. The uncontrollable of wind and the hostile environment of the unit lead to the frequent occurrence of wind turbine, which has seriously affect the reliability of the unit operating, and even caused catastrophic damage.Whether the unit can be safe and normal operation is an important factor to improve the reliability of the unit, fault diagnosis has become the focus in the study of wind power generation technology. Therefore, it is very necessary for large wind turbines to carry out fault diagnosis.To improve the reliability of wind turbine, the methods of fault diagnosis were considered to be one of the most important research content.The whole content were focused on the core components of wind turbine,such as the generator and bearing, through wavelet transform, and then studied the prediction of crack failure trend of bearing by experiments. The main research contents were as follows:1. The structure and familiar fault of the wind turbine were studied, and the failure mechanism was analyzed. The modal analysis of Ansys bearing had showed the range of the frequency when the fault occurred;2. The generator rotor broken bar fault signal extraction of the wind turbine was studied. According to the characteristics of weak and unstable signal in early stage of the wind power generator, the wavelet packet was selected. The generator rotor broken bar fault was used as an example to demonstrate the good analysis ability of wavelet packet for weak signal.3. The diagnosis method of the generator rotor broken bar were studied,and the stator current was extracted as the fault feature of the analysis object, and the frequency and energy distribution of the weak faults were established, the superiority of the wavelet packet energy analysis in the diagnosis of generator rotor broken bar fault was verified;4. The prediction of the crack fault trend of bearing was carried out, and the wavelet neural network was studied deeply, the accuracy of the prediction method was verified by an example analysis, and provides a theoretical basis for wind field maintenance crew.