Font Size: a A A

The Study Of Fault Feature Extraction Technology About Gearbox Of Wind Turbines

Posted on:2015-02-08Degree:MasterType:Thesis
Country:ChinaCandidate:Y ZhangFull Text:PDF
GTID:2272330431981636Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
With the development of wind power technology, the unit capacity of wind turbines is increasing, at the same time,it has much higher requirements of real-time fault diagnosis, accuracy and validity.Gear box in wind turbines is not only a power transmission part,but also a high failure rate component, so,in order to ensure the normal operation of the wind turbines,it is very important to research the fault diagnosis of wind turbine gearbox.However, on research of the fault diagnosis of wind turbine gearbox, it is the key to extract the fault feature of vibration signals, which directly affects the validity,accuracy of wind turbine gearbox fault diagnosis, and reliability of fault early warning.By analyzing the types and causes of failures in wind turbine gearboxs and the feature characteristics of different fault signals, this paper gives the mathematical models of the gearbox vibration signal,and the method of feature extraction about gear box in wind turbines.In wind turbines,to the non-stationary and time-varying characteristics of gearbox fault vibration signal as a starting point, first of all, in order to more efficiently extract its signal characteristic,the preprocessing method based on wavelet noise reduction is proposed,giving the algorithm of the wavelet noise reduction; Secondly, After in-depth researching the Hilbert transform and EMD decomposition technique,this paper proposes the envelope demodulation method based on improved Hilbert transform to achieve fault feature, which uses EMD technology to varying gearbox fault vibration signals into the linear combination of many non-stationary IMF (intrinsic mode Function), then performs its IMF components with Hilbert transform of envelope demodulation, obtains fault signal characteristics from the spectral analysis of its envelope signal; Finally, the numerical example validate that the method can highlight doubling phenomenon, reduce its fault errors and extract the fault characteristics of wind turbine gearboxes effectively, which is a more reasonable program about fault feature extraction,with high availability, accuracy, validity and reliability...
Keywords/Search Tags:wind turbine gearbox, wavelet noise reduction, Hilbert transform, EMD, fault feature extraction
PDF Full Text Request
Related items