【摘要】 集合经验模式分解(Ensemble Empirical Mode Decomposition,EEMD)存在着辅助白噪声难以消除和容易产生虚假模式的缺陷。针对EEMD方法在齿轮箱故障信号处理中的不足,将自适应噪声完备集合经验模式分解(Complete Ensemble Empirical Mode Decomposition with Adaptive Noise,CEEMDAN)应用于齿轮故障信号分析,提出了基于CEEMDAN能量熵的齿轮状态识别方法。该方法首先利用CEEMDAN分解齿轮振动信号,然后计算振动信号分解结果的能量熵,将能量熵作为特征参数来区分不同的齿轮运行状态。将该方法用于区分正常、轻度刮伤和中度刮伤齿轮运行状态,并与基于EMDEEMD能量熵的方法进行了对比。结果表明,该方法可以有效地区分相近的齿轮运行状态,与其他几种方法相比具有明显的优势。更多还原

【Abstract】 The Ensemble Empirical Mode Decomposition(EEMD) often encounters two difficulies in removing the added white noise residing in extracted components and easy production of spurious modes. Aiming at the deficiencies in the EEMD,the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN) is introduced to examine gearbox fault data and a method for condition identification of gearboxes based on CEEMDAN energy entropy is proposed. In the proposed method,the gearbox vibration signal is decomposed by using the CEEMDAN,then the energy entropy of the decomposition results is calculated and the energy entropy is taken as a characteristic parameter to identify different gear operating condition. Afterwards,the proposed method is used to discriminate between normal,slight-scratch and medium-scratch gear operating conditions,and compared with the method based on EMD \ EEMD energy entropy. The results show that the proposed method can effectively discriminate between these three similar gear operating conditions,and the proposed method has a clear advantage in condition identification of gearboxes.更多还原