基于循环相关和LPSO算法的自适应MCKD方法的滚动轴承早期故障特征提取

摘要

Aiming at the problem that the feature of incipient fault of rolling element bearing is difficult to be extracted under strong noise background,a fault diagnosis method based on MCKD and symmetrical differencing energy operator demodulation was proposed.The filter size L and the period of interesting signal T play an important role in MCKD filtering.An adaptive MCKD filter based on cyclic correlation and LPSO was proposed,which could search for the best parameters automatically.The feature after filtering was outstanding,but some residual noise was still remained.To reduce the residual noise and get the demodulation spectrum,symmetrical differencing energy operator demodulation was applied after filtering.The result of incipient fault diagnosis of rolling element bearing shows that the method is effective.%针对强噪声情况下滚动轴承早期故障信号特征难以提取的问题,提出了MCKD与对称差分能量算子解调的特征提取方法.MCKD算法进行滤波时,滤波器长度L和故障周期T对滤波效果的影响至关重要,因此提出基于循环相关和LPSO算法结合的自适应的MCKD算法,自动搜寻MCKD算法所需最优参数;原信号经滤波后,故障特征被明显突出,为了剔除剩余噪声,对滤波后信号进一步做对称差分能量算子解调,剔除剩余噪声同时获得解调谱,进而提取滚动轴承的早期故障.实验分析验证了该方法的有效性.