Application And Research Of Order Analysis In 1.5MW Wind Turbine Gearbox Fault Diagnosis

As a kind of renewable and no pollution new energy, wind energy source has been paid more and more attention to many countries, with the gradual increase in unit capacity and the wind turbine volume, the probability of the occurrence of the wind turbine is also increasing. As one of the most important of structure of the wind turbine, the gearbox failure and downtime caused by the frequency of its occurrence relative to other components are the highest. So it is very important to the real-time monitoring of wind turbine gearbox failures. At the same time, due to the dual effects of itself structure and wind speed, the collected wind turbine gearbox fault signals often have the character of non-stationary and contains a lot of noise. In order to extract the characteristic frequency of fault information, improving the efficiency of the operation, this paper will make from the following several aspects of wind turbine gearbox vibration signal analysis and fault diagnosis:First, according to the wind turbine gearboxes non-linear and non-stationary characteristics of the vibration signals, the structure and typical failure forms of gearbox is expounded in detail. On the basis of this, we make a further study about the frequency of the vibration signal characteristics and mechanization. And about the typical fault vibration signal has been analyzed which is collected by the scene, illustrates the characteristics of different fault vibration signal.Second, diagnosis method in time domain and frequency domain on the early fault diagnosis method of the wind turbine gearbox were studied, and the traditional FFT analysis method and principle of order analysis method based on the COT re-sampling made in-depth study, illustrates the traditional common spectrum analysis method based on FFT can only extract the characteristic frequency of machinery in a stable speed condition, if used to extract the fault characteristic frequency fan, it will be hard to identify the frequency of "fuzzy" phenomenon. Also used based on the analysis of COT re-sampling order processing non-stationary signal, eliminating the "fuzzy" frequency caused by rotating speed instability phenomenon, and the signal through the simulation compared with the traditional FFT analysis, validate the superiority of the method.Finally, as a result of the collected at the scene of the wind turbine gearbox contains a lot of noise signal, has affected the traditional order analysis in the process of the precision of the re-sampling, in this paper, the traditional order analysis to improve the re-sampling process, was proposed based on order analysis and wavelet threshold de-noising the diagnosis method of combining compared with heavy sampling order based on the COT, improve the accuracy of the re-sampling, and through the comparative analysis of the field case to verify the effectiveness of the proposed method.