摘要
针对轴承振动信号非线性、非平稳性和故障特征微弱性的特点,以及工程实际中难以获得大量故障样本的情况,提出了一种基于多尺度排列熵和支持向量机的轴承故障诊断新方法。该方法首先对轴承不同运行状态下的振动信号进行多尺度排列熵特征提取,然后通过距离评估技术从原始多尺度排列熵特征中选取敏感特征,最后将敏感特征输入到采用遗传算法优化的支持向量机中,实现对轴承不同运行状态的自动识别。对实验数据分析的结果表明,该方法可以精细地获取故障信息,从大量原始特征中选择出敏感特征,有效地实现滚动轴承故障状态的诊断。
Aiming at such thecharacteristics of bearing vibration signal as nonlinear,non-stationary and weakness of fault feature,and the situation that it is difficult to obtain a large number of fault samples in the practical engineering,a novelbearing fault diagnosis methodbased on multiscale permutation entropy and support vector machine isproposed. Firstly,multiscale permutation entropyfeatures are extracted from the bearing vibration signals under different running states. Secondly,with distance evaluation technique,the sensitive features are selected from the original multiscale permutation entropy features.Finally,the sensitive features are input into the support vector machineoptimized by genetic algorithm to automatically identify the different running states. The experimental results show that the proposed method can precisely extract fault information,select sensitive ones from a large number oforiginalfeatures,and effectively implement the diagnosis of rolling bearing fault state.
作者
瞿金秀
石长全
丁锋
王文娟
Qu Jinxiu;Shi Changquan;Ding Feng;Wang Wenjuan(Department of Mechanical and Electronic Engineering,Xi＇an Technological University,Xi＇an 710021,China;State Key Laboratoryfor Manufacturing System,Xi＇an Jiaotong University,Xi＇an 710049,China)
出处
《煤矿机械》
北大核心
2018年第9期143-146,共4页
Coal Mine Machinery
基金
陕西省自然科学基金青年项目(2017JQ5017)
关键词
多尺度排列熵
敏感特征选择
支持向量机
故障诊断
muhiscale permutation entropy
sensitive feature selection
support vector machine
fault diagnosis
