摘要
机车走行部滚动轴承的状况直接关系到机车的性能和列车的运行安全。针对目前机车走行部滚动轴承故障诊断准确率不高、模型构建时间较长的问题,提出一种基于小波包和贝叶斯分类的故障诊断方法。通过小波包变换构造故障特征集,利用粗糙集和主成分分析进行降维,将未降维和降维之后的故障特征集输入到贝叶斯分类模型中实现故障诊断,最后将贝叶斯分类方法和神经网络及最小二乘支持向量机方法进行比较。仿真结果表明,朴素贝叶斯分类方法构建模型的时间更短,分类准确率更高。
The status of the locomotive running gear rolling bearing is directly related to the locomotive perform-ance and the safe operation of the train.Aiming at solving such problem as low accuracy of the fault diagnosis and long model -construction time of locomotive running gear rolling bearing,this paper proposes a fault -diag-nosis method based on wavelet packets and Bayesian classification.The method need to construct fault feature set through wavelet packet transform,and make the use of rough set and principal component analysis to reduce the dimension,and then input the fault feature sets of dimension reduction before and after to the bayesian classifica-tion model to achieve fault diagnosis in turns,and finally make a comparison among the bayesian classification method and the neural network and least squares support vector machine method.The simulation results show that the time of building model with the method of naive bayes classification is shorter,and the classification ac-curacy is higher.
出处
《铁道科学与工程学报》
CAS
CSCD
北大核心
2015年第3期636-642,共7页
Journal of Railway Science and Engineering
基金
国家自然科学基金资助项目(51165001)
广西自然科学基金面上资助项目(2013GXNSFAA019297)
关键词
机车走行部
滚动轴承
故障诊断
小波包
贝叶斯分类
locomotive running gear
rolling bearing
fault diagnosis
wavelet packet
bayesian classification