基于补充总体局部均值分解的轴承故障诊断方法-《郑州大学学报(工学版)》

文章信息/Info

Title:: Resarch on Fault Diagnosis Method of Bearing Based on Complementary Ensemble Local Mean Decomposition

Author(s):: Ren Zihui; Qu Hu; Wang Cui; Chen Ming; School of Information and Control Engineering, China University of Mining and Technology, Xuzhou, Jiangsu, 221008

Keywords:: CELMD; characteristic frequency; FFT transform; vibration signal; roller bearing

摘要:: 为了弥补局部均值分解(LMD)在处理非平稳、非高斯信号的不足，提出一种基于补充总体局部均值分解(CELMD)和频谱分析相结合的轴承故障诊断方法。该方法向原信号成对地添加符号相反的白噪声，首先对含噪信号进行LMD分解，得到一系列的乘积函数(PF)，再选取包含最丰富故障信息的PF分量，最后对该PF分量进行FFT变换，提取故障特征频率，实现对轴承状态和故障类型的识别。通过对仿真信号和轴承振动信号的分析，表明该方法不仅能消除残留白噪声和抑制模态混叠还可以提高故障诊断的准确性和有效性。

Abstract:: To solve the problem that local mean decomposition (LMD) method was not insufficient in process the non stationary and non Gaussian signal, a fault diagnosis method based on the complementary ensemble local mean decomposition (CELMD) and spectrum analysis was proposed. Firstly, in this method,the white noises were added in pairs into a target signal, and then the noisy signal was decomposed into a series of production function by using LMD method.The PF component containing main fault information was selected, which was transformed by fast Fourier transform(FFT), to realize the identifications of the working status and fault types. Through the analysis of the simulation signals and the vibration signal of the bearing, it was proved that the method could eliminate the residual white noise and restrain the mode mixing, and improve the accuracy of the fault diagnosis as well.