Train localization by mutual correction of acceleration and interior sound
Train localization by mutual correction of acceleration and interior sound
复制标题
通过加速度和车内声音相互校正进行列车定位
DOI：
10.11532/structcivil.62a.571
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
蘇迪，佐野聡，田中博文，長山智則，水谷司
中科院分区：
文献类型：
--
作者：
M. Hori;Y. Ono;雨宮 智宏;蘇迪，佐野聡，田中博文，長山智則，水谷司
在这项研究中，列车定位的方法，通过相互校正的加速度和内部声音提出并实现与智能手机。首先，在行进方向上的加速度被积分以获得近似的列车速度。然后，通过从铁轨接头声音估计的速度来校正该速度。利用稀疏表示和多分辨率分析去除背景噪声，提取清晰的钢轨接头声。最后，通过对校正后的速度进行积分来识别列车位置。在实验列车上进行了验证试验。通过智能手机的实现，可以达到5%的识别准确率。这种方法有望通过解决一些悬而未决的问题，特别是在GPS信号不可用的情况下，为轨道管理提供一个简单而准确的列车定位工具。
In this study, a method of train localization through the mutual correction of acceleration and interior sound is proposed and implemented with smart phones. First the acceleration in the travel direction is integrated to obtain the approximate train speed. Then this speed is corrected by the speed estimated from the rail joint sound. The sparse representation and multi-resolution analysis can be utilized to remove the background noise to extract clear rail joint sound. Finally, the train location is identified by the integration of the corrected speed. A verification test has been carried out in the experiment train. The implement by smart phones can achieve a 5% identification accuracy. This approach is expected to provide a simple and accurate train localization tool for track management by resolving some outstanding issues, in particular where GPS signal is not available.