风电机组滚动轴承复合故障诊断研究

RESEARCH ON COMPOSITE FAULT DIAGNOSIS OF WIND TURBINE ROLLING BEARINGS

针对在强风电机组背景噪声下进行滚动轴承复合故障诊断时,由于故障之间的相互联系、交叉影响使得多种故障特征混叠在一起,易造成漏诊、误判等问题,提出一种基于多点最优调整的最小熵解卷积(multipoint optimal minimum entropy deconvolution adjusted,MOMEDA)与1.5维能量谱相结合的风电机组滚动轴承复合故障诊断方法;首先利用MOMEDA算法对原始滚动轴承振动信号进行解卷积预处理;然后对解卷积信号进行1.5维能量谱分析;最后通过分析谱图中幅值突出的频率成分来判断故障类型。仿真信号和应用实例分析结果表明,该方法能够有效提取出在强背景噪声下的复合故障特征,实现风电机组轴承复合故障的准确诊断。

Multiple fault features are easily mixed together,and can cause missed diagnosis and misjudged in the case of composite fault diagnosis of rolling bearing for wind turbines. The multi-point optimal minimum entropy deconvolution adjustment(MOMEDA)and 1.5-dimensional energy spectrum are combined to diagnose the composite faults of wind turbine rolling bearings. Firstly,the MOMEDA algorithm is used to deconvolute the original rolling bearing vibration signal. Then,the deconvoluted signal is processes by applying 1.5-dimensional energy spectrum;Finally,the fault type is judged by analyzing the frequency components whose amplitudes are prominent in the spectrum. The simulation signal and application example analysis show that the proposed method can effectively extract the composite fault characteristics under strong background noise and realize the accurate diagnosis of wind turbine bearing composite faults.