基于多目标优化MCKD的轴承微弱故障诊断研究

Research on Weak Fault Diagnosis of Bearing Based on Multi-objective Optimization MCKD

针对最大相关峭度反褶积(MCKD)算法中滤波步长对轴承微弱故障信号处理准确性的影响较为严重这一缺陷,提出基于多目标优化MCKD的自适应轴承微弱故障诊断方法。首先针对多尺度极差熵(MRE)对故障信号中的有用信息不能全面反映的缺陷,对其进行均值化处理形成多尺度均值极差熵(MMRE);随后以MMRE和符号动态熵(SDE)为目标函数,采用多目标粒子群优化(MOPSO)算法对MCKD算法中的滤波步长进行优化选取,使得MCKD算法可以自适应地处理轴承微弱故障信号。通过试验数据对该方法进行验证并与单目标优化法进行对比,证明了该方法的有效性及可靠性。

Aiming at the defect that the filtering step in the maximum correlation kurtosis deconvolution(MCKD)algorithm has a serious influence on the accuracy of the bearing weak fault signal processing,proposed an adaptive bearing weak fault diagnosis method based on multi-objective optimization of MCKD.Firstly,aiming at the defect that the useful information in the fault signals cannot be fully reflected by multi-scale extreme entrop(MRE),it was homogenized to form multi-scale mean mean extreme entrop(MMRE);then took MMRE and symbolic dynamic entrop(SDE)as the objective function,the multiobjective particle swarm optimizati(MOPSO)algorithm was used to optimize the selection of the filtering step in the MCKD algorithm,so that the MCKD algorithm can process the weak fault signals of the bearings in an adaptive way.This method was validated by experimental data and compared with the single-objective optimization method to prove the effectiveness and reliability of the method.