基于LFK和熵谱差分准则的滚动轴承信号预处理方法研究  

Study on the Preprocessing Method of Rolling Bearing Signal based on LFK and Entropy Difference Spectrum Criterion

作　　者：于贺[1] 李洪儒[1] 孙健[1] 许葆华[1] 

机构地区：[1]军械工程学院,河北石家庄050003

出　　处：《机械传动》2016年第12期32-37,共6页Journal of Mechanical Transmission

基　　金：国家自然科学基金项目"基于复杂度特征的旋转机械通用部件剩余寿命预测研究"(51541506)

摘　　要：针对滚动轴承故障信息易被不确定性随机噪声湮没的问题,提出了基于LFK的滚动轴承信号降噪预处理方法。首先,通过局部特征尺度分解对信号进行分解并提出熵值差分谱对分量进行筛选重构;然后,对重构信号中可能存在的残余噪声,采用快速峭度图算法进一步滤波降噪,一定程度上去除了噪声的干扰;同时也较好地保留了故障特征信息。最后,利用仿真信号和滚动轴承内圈故障信号验证了该方法的有效性。Aiming at the problem that the fault feature of rolling bearing can be easily overwhelmed by random noise, a novel denoising method on the basis of local characteristic - scale decomposition and Fast Kurtogram （LFK） is presented. Firstly, the signal is decomposed by local characteristic - scale decomposition （LCD） and entropy difference spectrum is proposed to choose and reconstruct the component signals. However, the residual noise may still exist in the reconstructed signal. So in order to remove noise interference to some degree, the Fast Kurtogram is applied to enhance the filtering and denoising effect further. Meanwhile, and the bearing fault feature information is well preserved. At last, the validity of the method is tested and verified by the simulation signal and the inner ring fault signal of rolling bearing.

关 键 词：局部特征尺度分解 熵值差分谱 快速峭度图 去噪 滚动轴承 

分 类 号：TH133.33[机械工程—机械制造及自动化]