【摘要】 针对滚动轴承早期故障特征难以从原始振动信号中提取的问题,提出了一种基于参数优化的变分模态分解(Variational Mode Decomposition, VMD)的轴承早期故障诊断方法。利用天牛须搜索算法(Beetle Antennae Search, BAS)对VMD算法的最佳参数组合进行优化搜索,搜索过程中以VMD分解后各本征模态函数(Intrinsic Mode Function, IMF)峭度值的倒数作为适应度函数。搜索结束后根据所得结果设定VMD算法的IMF分量个数和二次惩罚因子,并利用参数优化VMD算法对轴承振动信号进行分解。借助峭度准则筛选出最佳IMF分量进行Hilbert包络解调运算,获取信号的包络谱,包络谱中可显现出较为明显的故障冲击特征,根据这些冲击成分可实现轴承早期故障诊断。经过与经验模态分解(Empirical Mode Decomposition, EMD)和固定参数VMD算法的试验对比,所述方法可以更有效地提取轴承早期故障的特征。更多还原

【Abstract】 Aiming at the problem of incipient fault features being difficult to extract in original vibration signals of rolling bearing, an incipient fault diagnosis method of rolling bearing based on variational mode decomposition(VMD) with parameters optimized was proposed. Firstly, Beetle antennae search(BAS) algorithm was used to search the optimal parameter combination of VMD algorithm. The reciprocals of kurtosis values of intrinsic mode functions(IMFs) obtained with VMD were taken as fitness functions in the search process. The number of IMFs and the quadratic penalty factor of VMD algorithm were set up according to the obtained results after search. Then, the bearing vibration signal was decomposed using VMD algorithm with parameters optimized, and the optimal IMF component was chosen with the kurtosis criterion.Hilbert envelope demodulation calculation was done for the optimal IMF component to gain its envelope spectrum. This envelope spectrum could reveal more obvious fault impulse features to realize incipient fault diagnosis of rolling bearing. The results were compared with those obtained using EMD, VMD with fixed parameters and tests results showed that the proposed method can more effectively extract incipient fault features of rolling bearing.更多还原