Operational conditions division of wind turbines

Due to the complicated and variable operating environment of wind turbines, a method based on k-means clustering algorithm is proposed to divide the operating condition for wind turbines and supplies the forceful foundation for making a maintenance plan. In this method, the fault characteristic analysis and the characteristic frequency calculation for the gear box and the generator are proposed. And then, the vibration characteristics of the two subsystems are extracted by using the wavelet packet theory. Finally, the k-means clustering algorithm is used to classify the measured data from a wind farm. According to the central-limit theorem, the criterion is established to evaluate whether the state characteristic parameters are normal or not. It's beneficial to the early warning under the different working conditions of wind turbines. The results show that this division is accurate and it is beneficial to evaluate the operating states of wind turbines.