| As human society rapidly growing, energy demanded and climate degeneration, environment problem, energy problem becomes increasingly prominent. Wind power as one of the large-scale development and utilization of new renewable clean energy,in the technical research, equipment manufacturing and supporting parts, etc, have been developed rapidly. Recent years, in our country the wind power generation has been vigorously developed. With the rapid development of wind power equipment, it gradually towards large-scale. Then with the gradual development, many problems are gradually exposed about the large-scale wind turbine fault diagnosis and dynamic monitoring. For large-scale wind turbine, strong interference and weak feature signals, generation and development of equipment failure, the change of working conditions led to the non-stationary characteristics of vibration signals. Which makes wind turbines nonstationary vibration signal feature extraction has become an important research direction, widely focused on noise reduction preprocessing, instantaneous frequency of signal feature extraction, the determination of diagnosis is the main technology research questions.Wind turbine transmission system is one of the important components of the wind turbine. At present, the dynamics analysis of gear-box usually focused on the planetary gear train. Planetary gear train is also an important component in wind turbine transmission system. Based on the lumped mass method, the dynamics model of megawatt wind turbine gear transmission system is build. And analyze the mathematical expressions of the wind turbine gear transmission system while the vibration excitation and the vibration characteristics are changing with the working condition.For wind turbine drive system, the analysis of the signal is stationary signal, may contain the rush, mutation and so on, also may contain non stationary white noise, and these tend to be the high frequency part of signal, at the same time due to the weak signals of the fault information contained in the general, often submerged in the noise signal, to analyze this kind of signal, signal to noise reduction processing. Therefore, this article puts forward in signal noise reduction based on SVD and gauss wavelet filter noise reduction method. The method can effectively isolate useful vibration signal of wind turbine drive system, fully to avoid contamination of strong background noise to signal, for follow-up to state characteristic signal time-frequency analysis to ensure premise.In the method of signal analysis, the superior adaptive characteristics of the HHT method is widely recognized, but such as the endpoint effect and modal aliasing problems are existed, impeding the application in the field of fault feature extraction. This paper put forward and realized the improved HHT analysis method based on the improved masking signal method to extract the non-stationary vibration signal feature. By adding the masking signal effectively suppresses the modal aliasing phenomenon, and resolved the endpoint effect problem based on the original data according to continuation the masking signal.This paper developed a software for the wind turbine drive component vibration analysis system, in order to facilitate to monitor status of wind turbine transmission system, and control the data of state degradation about the wind turbine transmission system. Design the whole scheme of wind turbine transmission parts vibration analysis system, applying modern signal processing technology, wireless sensor technology, virtual instrument technology and database technology to develop the subsystem. Then measuring the real data of the one of the wind turbine transmission system in the wind field, the signal has been processed and analyzed. Finally, the reliability and effectiveness of the system has been verified. |
PDF Full Text Request