滚动轴承故障诊断的仿真研究

摘要

Research fault diagnosis jof rolling bearing. Fault vibration signals are of non-stationariness and muta-genicity, it is difficult using conventional Fourier-transform or single wavelet analysis to extract the signals character-istics accurately. In order to improve the accuracy of fault diagnosis of rolling bearing, the paper proposed a fault di-agnosis method of rolling bearing based on packet analysis and Hilbert transform. First, the packet analysis was used to decompose the collected rolling bearing signals and eliminate the noises. Then the Hilbert transform was used for the fine signal decomposition. The MATLAB software was used for rolling bearing fault simulation experiment, and the simulation results show that this algorithm can improve the accurate rate of the rolling bearing fault diagnosis, and is suitable for processing the fault signals of rolling bearing.%研究滚动轴承故障诊断问题,故障振动信号具有非平稳性、突变性.由于运行中噪声影响识别故障信号、传统傅立叶变换或单一小波分析难以对特征信号进行准确提取,导致滚动轴承故障诊断正确率较低.为了提高了滚动轴承故障诊断正确率,提出一种小波分析和Hilbert变换的滚动轴承故障诊断方法.首先采用小波分析对采集滚动轴承信号进行分解,消除噪声信息,然后采用Hilbert变换对信号进行进一步精细分解.利用MATLAB软件对滚动轴承故障进行仿真,仿真结果表明,改进算法提高了滚动轴承故障诊断正确率,很适合处理滚动轴承的故障信号.