【摘要】 针对旋转机械运行过程中，滚动轴承会产生冲击性强、结构复杂的振动信号，导致故障特征难以精准提取的问题，本文提出一种基于冠豪猪优化算法（CPO）联合优化ICEEMDAN与MCKD的故障特征提取方法．首先，采用CPO算法对ICEEMDAN中噪声幅值（Nstd）与噪声添加次数（NR）等关键参数进行自适应寻优，分解原始带噪信号，得到多个本征模态函数（IMF），并以峭度为指标筛选有效分量进行重构；其次，在MCKD方法中引入CPO优化算法对其核心参数滤波器长度（L）与移位数（M）进行寻优，对重构信号应用MCKD以增强其包含的周期性冲击特征；最后，从包络谱中提取关键故障特征频率．实验结果表明，该方法能够有效提取关键故障特征频率成分，将提取效果与其他算法进行对比分析，ICEEMDAN-MCKD算法提取效果更加清晰，验证了其在故障特征提取方面的鲁棒性．更多还原

【Abstract】 During the operation of rotating machinery, rolling bearings generate vibration signals that have strong impact and complex structures, which makes it difficult to accurately extract fault features. Therefore, a fault feature extraction method based on the crested porcupine optimizer(CPO) combined with the optimization of ICEEMDAN and MCKD is proposed. Firstly, the CPO algorithm is used to adaptively optimize the key parameters such as noise amplitude(Nstd) and addition times(NR) in ICEEMDAN, decomposing the original noisy signal to obtain multiple intrinsic mode function(IMF), and selecting the effective components as indicators of kurtosis for reconstruction; Secondly, the CPO optimization algorithm is introduced into the MCKD method to optimize its core parameters, filter length(L) and shift number(M), and apply MCKD to the reconstructed signal to enhance the periodic impact features contained therein; Finally, the key fault feature frequencies are extracted from the envelope spectrum. Experimental results show that this method can effectively extract the key fault feature frequency components. The extraction effect is compared and analyzed with other algorithms. The ICEEMDAN-MCKD algorithm has a clearer extraction effect, verifying its robustness in fault feature extraction.更多还原