复选降噪自适应型MCKD方法研究

针对强噪声干扰下，最大相关峭度解卷积（Maximum correlation kurtosis deconvolution，MCKD）对于弱响应轴承滚动体故障信号指定周期冲击增强和辨识能力有限，无法自适应确定参数的问题，提出一种改进MCKD故障诊断方法。首先利用小波多尺度分解得到故障响应高频分量使冲击成份更加凸显；然后以峭度值最大准则复选出最优故障信号高频分量，降低噪音的干扰；最后结合小波方差自适应确定MCKD参数。轴承故障仿真、实验数据分析结果表明，该方法能够实现弱响应的轴承滚动体故障诊断，同时适用轴承内外圈故障诊断。

Study on Adaptive MCKD Method for Noise Reduction by Reselection

In order to solve the problem that the maximum correlation kurtosis deconvolution (MCKD) is unable to adaptively determine parameters due to its limited ability to specify periodic impact enhancement for bearing rolling element fault signals with weak response under strong noise interference, an improved MCKD fault diagnosis method was proposed in this paper. First, the high frequency component of fault response is obtained by wavelet multi-scale decomposition to make the impact component more prominent; then the optimal high-frequency component of fault signal is reselected by the maximum kurtosis criterion to reduce the noise interference; finally, the MCKD parameters are determined by wavelet variance adaptive. Bearing fault simulation and experimental data analysis show that this method can realize the fault diagnosis of bearing rolling element with weak response, and is suitable for the fault diagnosis of bearing inner and outer ringsunder strong noise interference.