基于小波分形理论的风电轴承故障识别

针对风电轴承振动特征信号易被环境噪声调制污染、信噪比低、具有非线性和不平稳的特点,利用基于小波分形的故障识别方法对此进行了研究.采用小波包分解,利用互信息法和Cao算法分别确定了相空间的延迟时间和嵌入维数,根据不同频带的关联维数变化确定风电轴承的工作状态.该方法不依赖于风力机工作的动力学模型,对整体系统信息状态变化敏感.通过现场实验证明,该方法较好地解决了风电轴承故障难以识别的问题,为更加细致地研究风电轴承振动信号提供了重要参考.

Fault recognition of wind turbine bearings based on wavelet and fractal theory

In order to solve the problems that the vibration signals of wind turbine bearings are easily modulated and polluted by the environmental noises and have such characteristics as low signal to noise ratio, non linearity and non stationarity, the corresponding study was performed with a fault recognition method based on wavelet and fractal theory. Through adopting the wavelet packet decomposition, the delay time and embedding dimension of phase space were determined with the mutual information method and Cao algorithm, respectively. The working state of wind turbine bearings was determined according to the correlation dimension changes of different frequency bands. The proposed method is independent on the working dynamical model for wind turbine, and is sensitive to the information state change of overall system. With the field experiments, it is found that the proposed method can better solve the hard distinguishing problem in the faults of wind turbine bearings, and provide the important reference for more detailed study on the vibration signals of wind turbine bearings.