基于变分模态分解和混沌麻雀搜索算法优化支持向量机的滚动轴承故障诊断

Rolling bearing fault diagnosis based on variational mode decomposition,chaotic sparrow search algorithm and support vector machine

针对滚动轴承振动信号中的故障特征难以提取的问题,提出一种基于变分模态分解(VMD)和混沌麻雀搜索算法(CSSA)优化支持向量机(SVM)的故障诊断方法。首先,利用VMD处理滚动轴承信号,提取本征模态分量(IMF)的能量谱和能量熵作为故障特征向量;其次,通过引入改进Tent混沌映射和自适应t分布策略,加入边界探索和警戒解除机制,对麻雀搜索算法(SSA)进行改进;最后,采用CSSA-SVM模型进行滚动轴承故障的识别和诊断。实验结果表明,CSSA-SVM模型能够有效识别滚动轴承的故障类型,拥有更高的诊断精度。

Aiming at the difficulty in extracting fault features from vibration signals of rolling bearing,a fault diagnosis method based on Variational Mode Decomposition(VMD),Chaotic Sparrow Search Algorithm(CSSA)and Support Vector Machine(SVM)was proposed. Firstly,the signals of rolling bearing were processed by VMD,and the energy spectrum and energy entropy of the Intrinsic Mode Function(IMF)were extracted as the fault feature vectors. Secondly,the Sparrow Search Algorithm(SSA)was improved by introducing the improved Tent chaotic map and adaptive t distribution strategy,and adding the boundary exploration and alarm cancellation mechanism. Finally,the CSSA-SVM model was used to identify and diagnose the rolling bearing faults. Experimental results show that the CSSA-SVM model can effectively identify the fault types of rolling bearing and has higher diagnosis accuracy.