对滚动轴承的故障特点进行了分析.根据小波包变换能将信号按任意时频分辨率分解到不同频段的特性,提出了小波包分解和重构用于滚动轴承信号消噪处理的概念及其算法,建立了滚动轴承加速度信号实时测试系统.通过实验研究得出,用小渡包分解并通过重构可使分解后的数据长度不变,使分解层数不受限制,证明小波包分析方法应用于滚动轴承信号消噪的有效性,并取得了较好的效果.
This paper analyzes fault characteristics of rolling bearing. Based on wavelet packet analysis theory,the concept of signal de-noise in rolling bearing with wavelet packet decomposition and reconstruction and the algorithm of wavelet packet energy feature extraction are presented. A real-time detection system of the acceleration signal of rolling bearing vibration is provided. It is found by experiments that wavelet packet decomposition and reconstruction make the lengths of data unchangeable and make the decomposed layers unlimited. This method is proved to be effective. It is achieved excellent result in signal de-noise of rolling bearings by experiment example.