基于共振稀疏分解和松鼠优化算法的滚动轴承故障诊断

共振稀疏分解方法在滚动轴承故障诊断方面得到广泛应用,分解参数的选取对故障分离效果起决定性影响。为保证参数选择的准确性,提出基于松鼠算法的自适应共振稀疏分解多参数优化方法。以信号低共振分量峭度最大作为目标,使用松鼠算法同时优化共振稀疏分解的品质因子与权重系数;利用最优品质因子和权重系数对滚动轴承振动信号进行共振稀疏分解,得到高低共振分量;对低共振分量进行希尔伯特包络谱分析。通过仿真试验和应用实例证明,所提方法可以有效提取轴承的微弱故障信息,实现共振稀疏分解小波基函数库与耗散函数之间的最优匹配,具有较高的分离精度。

Rolling bearing fault diagnosis with a resonance-based sparse decomposition and squirrel optimization algorithm

The resonance-based sparse decomposition method has been widely used in fault diagnosis of rolling bearings.The selection of decomposition parameters has a decisive impact on the effect of fault separation.To ensure the accuracy of parameter selection,an adaptive resonance-based sparse decomposition method with multi-parameter optimization was proposed.Firstly,aiming at maximizing the kurtosis of low resonance component,the squirrel algorithm was used to optimize the quality factor and weight coefficient at the same time.Secondly,the vibration signal of a rolling bearing was decomposed by the optimal quality factor and weight coefficient to obtain the high resonance-based component and the low resonance component.Finally,the Hilbert envelope spectrum analysis was performed on the low resonance component.The simulation experiment and application examples show that the proposed method can effectively extract the weak fault information of the bearing,realize the optimal matching between the wavelet basis function library and the cost function,and has high separation accuracy.