机车轴承故障优化诊断仿真研究  

Simulation Research on Optimization Diagnosis of Locomotive Rolling Bearing Faults

作　　者：徐国权 罗倩[1] 郭鹏飞 XU Guo-quan;LUO Qian;GUO Peng-fei(College of Information and Communication Engineering,Beijing Information Science and Technology University,BeijinglO0101,China)

机构地区：[1]北京信息科技大学信息与通信工程学院,北京100101

出　　处：《计算机仿真》2018年第9期166-171,194,共7页Computer Simulation

基　　金：国家自然科学基金资助项目(61271198);北京市科技提升计划项目(5211624101)

摘　　要：为了解决传统机车滚动轴承故障诊断算法采用的特征集敏感度不够和分类器需要大量样本的问题,提出了数据的自相关函数的波动性这一新特征与传统特征集相结合并采用灰色关联分析进行故障诊断的方法。数据的自相关函数的波动性能够非常有效地反映出滚动轴承各种状态的特性,具有较高的敏感度,将其与传统特征集结合组成新的特征集,提高特征集的敏感度。实际中的故障数据较少,灰色关联分析方法在数据量较少的情况下依然有较高的诊断正确率,更适用于实际的故障诊断。针对实际数据容易被掩盖的问题,使用共振解调技术对数据进行预处理。仿真结果表明,在数据量较少的情况下,优化方法比传统方法有更好的诊断效果。In order to solve the problems that the sensitivity of the characteristic set is not enough and a large number of samples are needed for the traditional fault diagnosis algorithm of the rolling bearings, a new fault diagnosis method is proposed. The new feature with the volatility of the autocorrelation function of data is combined the traditional feature set, and the grey relational analysis is used to diagnose the faults. The volatility of the autocorrelation function of data effectively reflects the characteristics of various states ＂of rolling bearings and has higher sensitivity, which can improve the sensitivity of the feature set. The actual fault data are less, and the grey relational analysis method has higher diagnostic accuracy in the case of less data. Therefore, it is more suitable for practical fault diagnosis. To solve the problem that the real data are easy to be concealed, the data are processed by resonance demodulation technique. The simulation results show that the optimization method has better diagnosis effect than the traditional method in the case of small amount of data.

关 键 词：滚动轴承 自相关函数 波动性 灰色关联分析 共振解调 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]