大型风力发电机组齿轮劣化故障诊断研究

Research on Gear Degradation Fault Diagnosis of Large-Scale Wind Turbine

针对大型风力发电机组齿轮出现不同劣化故障时对应频率范围内能量会发生变化的特点,提出了利用经验模式分解(EMD)能量分布作为故障特征向量,灰色相似关联度作为故障模式识别算法的大型风力发电机齿轮劣化故障诊断方法。首先,对采集到的原始信号进行EMD分解,运用相关系数法对获得的本征模式函数(IMF)进行筛选,剔除无意义的IMF分量;然后计算有效IMF的能量及能量比,构造故障特征向量;最后,根据待识别状态特征向量和已知标准状态故障特征向量的灰色相似关联度大小判断齿轮劣化故障类型。通过实验对所提方法进行了验证,结果表明,该方法能有效用于大型风力发电机齿轮常见的劣化故障诊断。

Considering the facts that different frequency range energy will change while the large-scale wind turbine gear degradation fault appeared,a method of large-scale wind turbine gear degradation fault diagnosis is proposed,which used empirical mode decomposition(EMD)energy distribution as fault feature vectors and grey similar incidence as fault pattern recognition algorithm.Firstly,the method of EMD is used to decompose the original vibration signal,and the method of correlation coefficient is used to screen intrinsic mode functions(IMF),removing the meaningless IMFs;then,the energy and energy distribution of effective IMFs is calculated,structural fault feature vector is obtained.Finally,the degradation fault type of gear is judged according to the gray similar incidence between the characteristic vector of the state to be recognized and the known standard condition.The proposed method is verified by experiments,and the results show that the method can be used in the common degradation fault diagnosis of large-scale wind turbine gear effectively.