基于广义精细复合多尺度量子熵和核主成分分析的中介轴承故障诊断方法  

Inter-shaft bearing fault diagnosis method based on generalized refined composite multiscale quantum entropy and kernel principal component analysis

作　　者：田晶[1] 张羽薇 张凤玲[1] 艾辛平 高崇[1] TIAN Jing;ZHANG Yuwei;ZHANG Fengling;AI Xinping;GAO Chong(Key Laboratory of Advanced Measurement and Test Technology for Aviation Propulsion System,Shenyang Aerospace University,Shenyang 110136,China)

机构地区：[1]沈阳航空航天大学辽宁省航空推进系统先进测试技术重点实验室,沈阳110136

出　　处：《航空动力学报》2024年第2期76-85,共10页Journal of Aerospace Power

基　　金：国家自然科学基金(12172231);沈阳市中青年科技创新人才支持计划(RC220439)。

摘　　要：针对中介轴承振动信号传递到机匣测表面上路径复杂,导致故障特征提取及识别困难等问题,提出了一种基于广义精细复合多尺度量子熵(generalized refined composite multiscale quantum entropy,GRCMQE)、核主成分分析(kernel principal component analysis,KPCA)与参数优化支持向量机的中介轴承故障诊断方法。该方法首先采用GRCMQE从振动信号中提取故障特征,构建高维故障特征集。其次,采用KPCA方法对高维特征数据降维,得到低维流形特征。然后,将得到的特征输入到基于交叉验证优化的支持向量机(cross validation-support vector machine,CV-SVM)中,完成故障模式识别。最后,在中介轴承故障数据集上对所提出的方法进行测试,结果表明该方法能够有效实现中介轴承不同故障类型的识别,并且故障识别精度达到98.33%。In view of the problems of complex paths of inter-shaft bearing vibration signal transmission to the measurement surface of the magazine,which lead to difficulties in fault feature extraction and identification,a fault diagnosis method based on generalized refined composite multiscale quantum entropy(GRCMQE),kernel principal component analysis(KPCA)and parameter optimization support vector machine was proposed for inter-shaft bearing fault diagnosis.Firstly,GRCMQE was used to extract fault features from vibration signals,and high-dimensional fault feature sets were constructed.Secondly,KPCA method was utilized to reduce the dimension of high-dimensional feature data to obtain low dimensional manifold features.Then,the obtained features were input into the support vector machine based on cross validation to complete the fault pattern recognition.Finally,the proposed method was tested on the intermediate bearing fault data set,and the results showed that the method can effectively identify different fault types of intermediate bearing,with the fault identification accuracy up to 98.33%.

关 键 词：中介轴承 振动信号 广义精细复合多尺度量子熵 核主成分分析 故障诊断 

分 类 号：V231.92[航空宇航科学与技术—航空宇航推进理论与工程]