基于EEMD和CICA的风电机组轴承故障特征提取

Bearing Fault Feature Extraction of Wind Turbine Based on EEMD and CICA

针对风电机组轴承故障振动信号具有高噪声、非线性、非平稳的特性,提出一种集成经验模态分解(ensemble empirical mode decomposition,EEMD)和约束独立分量分析(constrained independent component analysis,CICA)的故障特征提取方法(EEMD-CICA)。首先对采集的轴承振动信号进行预处理,消除噪声信号的影响,并利用EEMD进行分解,得到一系列本征模态函数(intrinsic mode function,IMF);然后利用邻近奇异值插值法估计源信号个数,并根据互信息准则选取IMF分量作为参考信号;最后根据选取的参考信号,利用CICA方法提取轴承故障中包含的信号特征,并进行包络解调得到包络谱。算例分析表明,此方法能有效提取轴承故障特征。

Taking into account the strong noise, nonlinear and non-stationary characteristics of bearing fault vibration signal of wind turbine,a fault extraction method based on ensemble empirical mode decomposition（EEMD） and constrained independent component analysis （CICA） is proposed.Firstly,the bearing vibration signal is pretreated to eliminate the influence of the noise signal and decomposed by EEMD to obtain a series of intrinsic mode functions （intrinsic mode function,IMF）.Then,the number of source signals in mixed signal is estimated by the adjacent singular value difference method.Besides, according to the mutual information criterion,several independent IMF components with better independence are chosen as the reference signals.Finafly,the CICA method is used to extract the vibration signal characteristics contained in the bearing fault based on the established reference signals, and envelope spectrum is obtained by envelope demodulation.The example analysis shows that this method can effectively extract bearing fault characteristics.