• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2017, Vol. 53 ›› Issue (8): 201-208.doi: 10.3901/JME.2017.08.201

• 制造工艺与装备 • 上一篇    

基于数学形态学和IFOA-SVR的滚动轴承 可靠度预测方法

康守强1, 叶立强1, 王玉静1, 谢金宝1,MIKULOVICH V I2   

  1. 1. 哈尔滨理工大学电气与电子工程学院 哈尔滨 150080;
    2. 白俄罗斯国立大学 明斯克 220030 白俄罗斯
  • 出版日期:2017-04-15 发布日期:2017-04-15
  • 作者简介:康守强(通信作者),男,1980年出生,博士,教授。主要研究方向为非平稳信号处理,故障诊断、状态评估与预测技术。E-mail:kangshouqiang@163.com
  • 基金资助:
    * 国家自然科学基金(51305109)、黑龙江省青年科学基金(QC2014C075)和哈尔滨理工大学青年拔尖创新人才(201511)资助项目; 20160419收到初稿,20170104收到修改稿;

Reliability Prediction Method of a Rolling Bearing Based on Mathematical Morphology and IFOA-SVR

KANG Shouqiang1, YE Liqiang1, WANG Yujing1, XIE Jinbao1, MIKULOVICH V I2   

  1. 1. School of Electrical and Electronic Engineering, Harbin University of Science and Technology, Harbin 150080;
    2. Belarusian State University, Minsk 220030, Belarus
  • Online:2017-04-15 Published:2017-04-15

摘要:

为了保证滚动轴承运行状态可靠度的预测精度同时增加预测步长,提出一种数学形态学分形维数结合改进果蝇优化算法-支持向量回归(Improved fruit fly optimization algorithm-support vector regression, IFOA-SVR)的滚动轴承可靠度预测方法。提取振动信号的包络信号,计算该包络信号的数学形态学分形维数,将其作为滚动轴承性能退化状态特征。利用IFOA对SVR中的参数C、g以及ε同时进行寻优,建立IFOA-SVR预测模型。利用极大似然估计结合IFOA建立威布尔比例故障率模型(Weibull proportional hazard model, WPHM),进而得到可靠度模型。将退化状态特征作为IFOA-SVR预测模型的输入,采用长期迭代预测法获取特征预测结果,并将该结果嵌入到可靠度模型中,从而预测出轴承运行状态的可靠度。试验结果表明,利用所提方法对滚动轴承可靠度进行预测,能在保证预测精度的前提下增加预测步长。

关键词: 果蝇优化算法, 可靠度预测, 数学形态学, 支持向量回归, 滚动轴承

Abstract:

In order to ensure the accuracy of the reliability prediction of a rolling bearing and increase the prediction step length, a rolling bearing reliability prediction method is proposed based on the fractal dimension of mathematical morphology and improved fruit fly optimization algorithm - support vector regression (IFOA-SVR). The envelope signal of the vibration signal is extracted and the fractal dimension of mathematical morphology of the envelope signal is calculated which is regarded as the performance degradation state feature of the rolling bearing. The IFOA is used to optimize the parametersC, g andε of SVR simultaneously, the IFOA-SVR prediction model is established. At the same time, the Weibull proportional hazard model (WPHM) can be established using the maximum likelihood estimation combined with IFOA, then the reliability model can be obtained. The performance degradation state feature is regarded as the input of the IFOA-SVR prediction model, the long-term iterative prediction method is used to obtain the prediction results of the feature, and the results are embedded in the reliability model, then the reliability of the rolling bearing running state can be predicted. Experimental results show that the proposed method can be used for the reliability prediction of a rolling bearing, and the prediction step length can be increased on the premise that the prediction accuracy is high.

Key words: fruit fly optimization algorithm, mathematical morphology, reliability prediction, support vector regression, rolling bearing