首页> 外文会议>IIE Annual Conference Exposition >Gearbox failure detection based on multivariate modeling and SPC
【24h】

Gearbox failure detection based on multivariate modeling and SPC

机译:基于多变量建模和SPC的变速箱故障检测

获取原文

摘要

This paper proposes a scheme for failure detection of a gearbox subject to vibration monitoring using multivariate autoregressive modeling and multivariate statistical process control (SPC). In this study, a wavelet method is used to extract the health indicators from the original vibration data representing health condition of the monitored machine. Vector Autoregressive (VAR) model is fitted to the multivariate health indicators data in order to obtain the VAR model residuals defined as the difference between the observations and the one-step-ahead predictions. Finally, the control chart for identifying the abnormal conditions is built based on the obtained VAR model residuals. The results show that this method can clearly detect the incipient gearbox failure. When the chart was applied to historical data, a clear pattern appeared on the chart, tracking failure propagation from the initial state to the full failure. This study demonstrates usefulness of the proposed methodology for the real application and the scheme can be applied also to other areas.
机译:本文提出了一种使用多元归共建模和多变量统计过程控制(SPC)对振动监测进行振动监测的齿轮箱的故障检测方案。在该研究中,小波方法用于从代表监控机器的健康状况的原始振动数据中提取健康指标。传染媒介自回归(VAR)模型适合多元健康指标数据,以便获得定义为观察与一步预测之间的差异的VAR模型残差。最后,基于所获得的VAR模型残差来构建用于识别异常情况的控制图。结果表明,该方法可以清楚地检测初始齿轮箱故障。当图表应用于历史数据时,图表上出现明确的模式,跟踪从初始状态到完全失败的故障传播。本研究表明了真实应用方法的有用性,并且该方案也可以应用于其他地区。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号