...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>IFAC PapersOnLine >On the Problem of On–Board Early Detection of Hunting on Rail Vehicles: an exploratory study
【24h】

On the Problem of On–Board Early Detection of Hunting on Rail Vehicles: an exploratory study

机译：关于轨道车辆岗位拆迁问题的问题：探索性研究

获取原文
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Hunting instability is a serious concern for high–speed trains, thus its early on–board detection is important for ensuring safety, passenger comfort, and avoiding unnecessary wear on the vehicle and the track. In this study two on–board hunting detection methods, employing the bogie lateral random vibration signal, are postulated and explored. The first is founded upon Information Entropy (IE) concepts, while the latter upon a measure of the predictive ability of an adaptive Recursive AutoRegressive (RAR) signal model. Their performance is systematically assessed via extensive Monte Carlo simulations using a SIMPACK–based high– fidelity vehicle model and three metrics: True Positive Rate (TPR), that is the rate of correct detection of hunting, False Positive Rate (FPR), that is false detection of hunting, and Detection Delay Time (DDT) with respect to hunting conventional onset. Interesting comparisons with a classical Root Mean Square (RMS) based method are also presented. The results demonstrate highest effectiveness for the RAR–based method, achieving a TPR of 100% @ 4% FPR and sample Mean Detection Delay Time 1.18 s, with the IE–based and the RMS–based methods following. A main problem with the classical RMS–based method relates to its high Detection Delay Time (sample mean DDT 6.50 s).
机译：狩猎不稳定是对高速列车的严重关注，因此其早期的载体检测对于确保安全性，乘客舒适性并避免在车辆和轨道上不必要的磨损是重要的。在这项研究中，采用转向架横向随机振动信号的两个板载狩猎检测方法被假设和探索。首先是在信息熵（IE）概念时，而后者在测量自适应递归自回归（RAR）信号模型的预测能力时。它们的性能通过广泛的蒙特卡罗模拟系统地评估了Simpack的高保真性车型和三个指标：真正的阳性率（TPR），即正确检测狩猎，假阳性率（FPR），即捕捉捕捉的假检测和检测延迟时间（DDT）捕捉常规发作。还提出了具有基于经典的均方根（RMS）方法的有趣比较。结果表明了基于RAR的方法的最高效果，实现了100％@ 4％FPR和样本平均检测延迟时间1.18S的TPR，以及基于IE的基于RMS的方法。基于经典RMS的方法的主要问题涉及其高检测延迟时间（样本均值DDT 6.50 S）。

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