滚动轴承出厂检测与智能故障诊断研究

Research on production test and intelligent diagnosis of rolling element bearing

由于传统滚动轴承出厂检测存在自动化程度低、劳动强度大、参数单一和临界判断模糊等缺点,本文结合滚动轴承振动原理,提出了数学形态学与KL距离相结合的滚动轴承故障检测方法.通过对滚动轴承信号进行多尺度形态滤波,定量分析形态算法,选取最优结构元素尺度;为进一步对滚动轴承信号进行故障诊断,对获取的滤波信号进行KL距离计算,实现轴承故障的检测.测试实例表明,形态膨胀滤波器能够有效抑制轴承故障中的噪声,并且可以有效地突出特征故障频率;数学形态学与KL距离相结合的滚动轴承故障检测方法具有较好的检测效率和精度.

Due to the defect of traditional rolling element bearing of production test,such as low degree of automation,high strengh of work,single parameter and blur of critical judgment,a fault detection method based on mathematical morphology and Kullback Leibler Divergence is introduced.The optimal structuring elements are crealed using multi-scale morphology algorithm;In order to recognize fault pattern of rolling element bearing,Kullback Leibler(KL)divergence is used to measure filter signals.Experiment results show that:firstly,morphology dilation filter can effectively suppress the noise and highlight characteristic fault frequency;secondly,the fault detection method has good efficiency and accuracy.