LMD与优化OMP算法的滚动轴承故障诊断方法研究

Study on Fault Diagnosis Method of Rolling Bearing Based on LMD and Optimized OMP Algorithm

针对K-SVD算法在构建字典时字典原子易受噪声干扰混入虚假原子,正交匹配追踪(OMP)算法不易区分相似原子以及迭代终止条件难以确定问题,提出一种基于局部均值分解(LMD)与优化OMP算法的故障特征提取模型。该模型首先采用LMD算法分解信号,根据皮尔逊相关系数法选取最优PF分量作为样本信号构建字典;然后在OMP算法基础上引入Jaccard系数和峭度最大准则,解出稀疏系数并重构信号;最后进行仿真和实验数据分析。结果表明,基于LMD与优化OMP的故障特征提取模型对滚动轴承故障特征提取效果有改善。

Aiming at the fact that dictionary atoms in the K-SVD algorithm are susceptible to noise and mixed with false atoms when constructing a dictionary,the Orthogonal Matching Pursuit(OMP)algorithm is difficult to distinguish similar atoms and the iteration termination condition is difficult to determine,a new fault feature extraction model based on local mean decomposition and OMP algorithm is proposed.The model first uses the LMD algorithm to decompose the signals,and selects the optimal PF component as a sample signal to construct a dictionary according to the Pearson correlation coefficient method.Then based on the OMP algorithm,Jaccard coefficient and kurtosis maximum criterion are introduced to solve the sparse coefficient and reconstruct the signal.Finally,simulation and experimental data analysis is performed.The results show that the fault feature extraction model based on LMD and optimized OMP has improved the fault feature extraction effect of rolling bearings.