改进的噪声总体集合经验模式分解方法在轴承故障诊断中的应用

在复杂的流程工业中,机械设备往往处在高速、重载、高温、高辐射的环境中,轴承作为主要的机械零部件起着重要作用。由于轴承故障振动信号的微弱和不平稳的特性,造成故障特征向量提取和故障诊断存在着困难。提出一种改进的CEEMDAN(Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise,ICEEMDAN)轴承故障诊断方法。通过对比分析仿真信号和实测信号可以得知:ICEEMDAN方法可以改善信号重构质量,具有良好的自适应性,能够提高故障信号的信噪比,从而可以有效地识别并提取有用的故障特征信息。

Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise and its Application in Bearing Fault Diagnosis

In the complex process of industry,the machinery equipment is often working in rigorous conditions such as highspeed,heavy load,high temperature and high radiation. Thus,the rolling bearing plays an important role in the maintaining mechanical components. To date,it is difficulty to extract the fault feature due to that the characteristics about the fault vibration signal is weak and unstable. The presents an improved CEEMDAN(ICEEMDAN)bearing fault diagnosis method.Researching by the simulation signal and the measured signal analysis,the proposed method can effectively improve the quality of reconstructed signals and have a good self-adaptability to enhance signal-to-noise of the fault signal. Therefore,it can identify and extract useful fault characteristic information effectively.