| Health monitoring of wind turbine blades is effective means to increase the service life of blade, reduce the maintenance costs of the blades and improve the efficiency of power generation. So it is very important to find the damage of blades timely without stop the wind turbine.This paper firstly introduces the method and the advantages and disadvantages of each method for the monitoring of the wind turbine blade damage at home and abroad. The basic structure and working principle of the wind turbine blade are expounded, and the common damage of the blade is classified.In this paper, the vibration characteristic of the blade is analyzed, when damages were appeared, the vibration characteristic will change obviously. The monitoring technology of the time frequency characteristic parameter based on the dynamic strain on the blade surface is discussed. The experiment device and the method of extracting the eigenvalue are introduced. Measure the strain distribution of the blade surface by fiber Bragg grating(FBG). Analysis the characteristics of the strain signal in time and frequency domain. Study a signal processing method based on dual tree complex wavelet transform(DT CWT). By comparing the characteristic parameter of the blades are in good condition and in the presence of damage, the working condition of the blade is judged.In this paper, a kind of non contact monitoring technology based on audio signal is studied. The noise of wind generator and microphone are introduced. Design and refit a microphone matched with the characteristics of the noise, introduce the experimental scheme. Collect the audio signal of the blades which are in good condition and in the presence of damage by the microphone, indicate the size of blade damage by using power spectrum, autocorrelation function and A weight 1/3 frequency doubling.Finally, collect the sound signal of 2.5MW wind turbine in Yancheng Dafeng wind field. Analysis the surrounding environment of the wind field and measurement methods. The experimental results show that the variation of the characteristic parameters of the audio signal generated during the monitoring of the wind turbine can be used to describe the damage degree of the blade. |
