基于EEMD和Hilbert分析的滚动轴承故障诊断方法研究

Research on Fault Diagnosis Method of Rolling Bearing Based on EEMD and Hilbert

滚动轴承在风电机组中广泛应用,其运行状态直接影响整台风机的性能。提出EEMD(总体平均经验模态分解)和Hilbert包络分析相结合的方法对滚动轴承进行故障诊断。经验模态分解具有自适应性,但存在一些不足,易产生虚假分量和模态混叠现象。针对EMD分解方法的不足,引入改进型算法EEMD。首先将振动加速度信号进行EEMD分解,计算各阶IMF峭度值的大小,选择峭度值较大的IMF分量,利用Hilbert变换对其进行包络谱分析,提取故障特征频率,辨识滚动轴承故障。通过对实验采集的滚动轴承振动信号进行分析,证明了该方法的有效性和准确性。

Rolling bearing is widely used in wind turbine scroll , the state of it will directly affect the performance of the whole typhoon machine. This paper presents a method of ＇ensemble empirical mode decomposition combined with Hilbert envelopment analysis to diagnose the fault diagnosis of rolling bearing. Empirical mode decomposition is adaptive, but there are some shortcomings, it is easy to produce false component and the mode mixing phenome- non. Aiming at the shortcomings of EMD decomposition method, EEMD algorithm is introduced in this paper. Firstly, the vibration acceleration signal is decomposed by EEMD, then by calculating each order IMF kurtosis value, selectting the IMF components which have the larger kurtosis values, then using Hilbert transform to do envelope spectrum analysis of them, we extract fault characteristic frequency to identify the rolling bearing fault. Through the analysis of the collected vibration signals of rolling bearings, it proves that the method is validity and accuracy.