基于全矢CEEMDAN能量矩和AMHSSA-SVM的滚动轴承故障诊断

Fault diagnosis of rolling bearings based on full vector CEEMDAN energy moment and AMHSSA-SVM

为充分利用滚动轴承的故障特征信息,提高故障诊断的准确性和可靠性,文中提出了一种基于全矢自适应噪声完全集成经验模态分解(Complementary Ensemble Empirical Mode Decomposition with Adaptive Noise,CEEMDAN)能量矩和自适应多种群混合麻雀搜索算法(Adaptive Multi-population Hybrid Sparrow Search Algorithm,AMHSSA)优化支持向量机(Support Vector Machine,SVM)的故障诊断方法。首先,采用全矢谱技术融合同源双通道信号;其次,采用CEEMDAN算法处理融合信号,选择相关系数较大的前5阶IMF分量,并计算其能量矩作为支持向量机模型的特征输入;最后,提出AMHSSA算法并优化支持向量机模型的参数,建立AMHSSA-SVM故障诊断模型。对该模型进行测试,结果表明:此模型有效提高了识别准确性,与类似模型对比,进一步证明了其在分类精度和优化时间方面的优越性。

In order to make full use of the fault characteristic information of rolling bearings and improve the accuracy and reliability of fault diagnosis,A Complementary Ensemble Empirical Mode Decomposition with Adaptive Noise,complementary ensemble empirical mode decomposition with adaptive noise,CEEMDAN Energy Moment and Adaptive Multi-population Hybrid Sparrow Search Algorithm,AMHSSA optimizes fault diagnosis methods of Support Vector Machine(SVM).First,homologous two-channel signals are fused by full vector spectrum technique.Secondly,the CEEMDAN algorithm is used to process the fusion signal,and the first 5 IMF components with large correlation coefficients are selected,and their energy moments are calculated as the feature inputs of the SVM model.Finally,AMHSSA algorithm is proposed and parameters of support vector machine model are optimized,and AMHSSA-SVM fault diagnosis model is established.The test results show that this model can effectively improve the recognition accuracy.Compared with similar models,this model further proves its superiority in classification accuracy and optimization time.