基于凸优化的RPSEMD及其在滚动轴承故障诊断中的应用

RPSEMD Based on Convex Optimization and Its Application in Rolling Bearing Fault Diagnosis

为改善再生相移正弦辅助经验模态分解(RPSEMD)在噪声影响下鲁棒性较差的缺陷,引入了一种广义的极小极大凹罚函数(GMC)作为1范数的替代,建立起了基于凸优化的降噪框架。将该凸优化降噪方法作为一种前处理手段,随后利用RPSEMD对预处理过的信号进行模态分解。数值仿真信号和实测轴承故障信号的试验结果,以及与EMD及EEMD的对比分析表明,该方法能够消除模态混叠现象的影响,有效提取轴承的故障特征频率。

To improve poor robustness of Regenerated Phase-Shifted Sinusoid Assisted EMD(RPSEMD)under influence of noise,a Generalized Minimax-Concave(GMC)penalty function is introduced as an alternative to l1 norm.A denoising framework is established based on convex optimization.The convex optimization denoising method is used as a preprocessing approach,and then the mode decomposition is carried out for preprocessed signals by RPSEMD.The numerical simulation signal and actual measured bearing fault signal and comparison analysis between EMD and EEMD show that the method eliminates influence of modal chaos phenomenon and effectively extract fault characteristic frequency of bearings.