Research On Fault Diagnosis Of Gearbox In Wind Turbines Based On Vibration Signal

As the large-scale-development clean energy, wind power in the world has a rapid development. Wind farms are usually built in remote open space, the work environment is relatively poor, as an important component of the wind turbine, internal structure of gearbox is complex and load is large, so it is a fault-prone area. Health gearbox can directly affect the economy and security of the entire plant operation, running condition monitoring and fault diagnosis of gearbox in wind turbines is very important.In order to realize the wind farm gear operating status of real-time monitoring, it is necessary to understand the internal structure of the gearbox and the operating principle, this paper analyzes the structure of the wind turbine currently two commonly-used gearbox, gearbox vibration model was established, combining typical vibration characteristic parameters gearbox failure analysis methods gearbox running real-time monitoring and fault diagnosis.Gearbox working environment is more severe, collected vibration signals often contain a lot of interference and noise components, in order to improve the reliability of the gearbox fault diagnosis in the vibration signal analysis collected before, generally must first noise reduction, improving signal to noise ratio to be analyzed. For less traditional wavelet noise, this paper introduces the second generation has a self-adaptive wavelet, and the simulation and the actual vibration data to verify the good performance in the second generation wavelet gearbox vibration noise in the signal.To improve the practical value, the paper discusses a common gearbox can basically cover the typical type of fault and reliable signal analysis methodology. The paper selected two typical failure data simulation experiments show is obtained by spectral analysis of vibration signals fault frequency components, by cepstral analysis and envelopment analysis enables the fault message clearer, more modulation information. To overcome the traditional methods of linear stationary signals are too dependent on the limitations, and to be able to simultaneously reflect signals when frequency characteristics, this article will apply HHT analysis to gearbox fault diagnosis, get a good diagnosis effect, more variety analytical methods complement each other, confirms another form stable and practical analytical methodology.Finally, in order to reflect the above analysis method of practicability, to apply it to a group of wind farm real analysis of vibration data through noise reduction processing, spectrum, cepstrum, after comprehensive analysis and comparison of the envelope spectrum and methods of HHT, vibration characteristics typical combination of gearbox failure, high speed pinion obtained partial failure of the conclusions, and field box to check the results match.