Application Of The Wavelet Analysis In Fault Feature Extraction Of Gearbox In Wind Turbine

With the rapid development of China wind power industry in recent years, the total installed capacity has reached75.32GW by the year of2012, occupying the first place in the global wind power installed capacity. But frequent wind turbine failures brings about a huge operation maintenance cost. The wind turbine gearbox has a high failure rate of the wind generator components, and the repair or the replacement is difficult. Therefore, the research to the fault eigenvalue extraction and the fault diagnosis analysis method for gearbox of the wind generator has giant economic value and realistic significance.Wavelet analysis method is a rapidly developing technology in recent years, which is widely used in signal analysis, image processing and communication engineering fields. This research puts the gearbox of the wind generator as the object of study, carrying on the theoretical research and case analysis of the gearbox fault eigenvalue extraction by means of wavelet analysis.Firstly,this article implements a basic research to the wavelet theory including the continuous wavelet transform, the discrete wavelet transform, the wavelet packet,and then uses the simulation example to explain its engineering. Secondly, the wavelet packet energy method and the wavelet denoising methods are studied, and the effectiveness of the method to the analysis of actual signal is verified. Thirdly, this paper studies a new signal processing algorithm based on the complex wavelet, termed multi-scale enveloping spectrogram. The new technique enables simultaneous multi-scale decomposition to extract and separate envelopes of the mechanical failure signals and overcomes the shortcoming of the conventional enveloping spectral analysis technique in which the bandwidth of the signal components of interest needs to be known in advance. Combining band-pass filtering and enveloping into a single-step operation and improving the robustness in signal analysis. The new technique is experimentally evaluated by using the vibration signals measured on the gear box of the wind turbine, verified the validity and effectiveness of the new algorithm. Finally, considering the the limitations of the first generation wavelet transform, this paper discusses the method to construct the second generation wavelet, and writes the program structure of the second generation wavelet through the Matlab software, using in the analysis of the actual signal the wind generator, proving the feasibility of the second generation wavelet.As the rapidly developed technology in the emerging field in recent years, the wavelet analysis has a deep theory and wide application, getting a very good application in the fault eigenvalue extraction of gearbox of wind turbines,and it has the very good application prospect in the field of other mechanical fault diagnosis. Therefore the wavelet theory is a technology worthing promoting.