基于PELT与稳健估计的高速铁路曲线控制点识别方法

谢智峰; 魏晖; 朱洪涛; 杨飞; 吴仕凤; 赵琳渊; 李昌

doi:10.13238/j.issn.1004-2954.202406190007

English Version

您当前的位置：

首页 >

文章列表页 >

基于PELT与稳健估计的高速铁路曲线控制点识别方法

线路/路基/轨道 | 更新时间：2026-06-22

- 基于PELT与稳健估计的高速铁路曲线控制点识别方法
- High-Speed Railway Curve Control Point Identification Method Based on PELT and Robust Estimation
- 铁道标准设计 2026年70卷第4期页码：40-47
- 作者机构：
  
  江西科技学院协同创新中心，南昌　330098
  江西省铁路大数据技术开发与应用工程研究中心，南昌　330098
  江西日月明测控科技股份有限公司，南昌　330096
  中国铁道科学研究院集团有限公司基础设施检测研究所，北京　100081
  中国铁路南昌局集团有限公司南昌高铁基础设施段，南昌　330103
- 作者简介：
  
  作者简介：谢智峰（1993—），男，讲师，2019年毕业于南昌工程学院动力工程专业，工学硕士，主要从事数据挖掘与算法优化研究工作，E-mail:xiezhifeng jxut@163. com。
  通信作者：魏晖（1975—），男，教授，工学博士，主要从事轨道平顺性表征与控制研究工作，E-mail:weihui@163. com。
- 基金信息：
  
  国家重点研发计划课题(2022YFB2602905);国家自然科学基金项目(52278465;52068052);江西省自然科学基金项目(20202BABL204056);江西省重大科技研发专项(20203ABC28W008);江西省汽车服务工程及产业升级协同创新中心开放基金(232KFJJ04)
- DOI：10.13238/j.issn.1004-2954.202406190007
  中图分类号： U238;U216.3
- 收稿：2024-06-19，
  
  修回：2024-07-26，
  
  纸质出版：2026-04-20
- 稿件说明：
移动端阅览
谢智峰, 魏晖, 朱洪涛, 等. 基于PELT与稳健估计的高速铁路曲线控制点识别方法[J]. 铁道标准设计, 2026,70(4):40-47.

XIE Zhifeng, WEI Hui, ZHU Hongtao, et al. High-Speed Railway Curve Control Point Identification Method Based on PELT and Robust Estimation[J]. Railway Standard Design, 2026, 70(4): 40-47.
谢智峰, 魏晖, 朱洪涛, 等. 基于PELT与稳健估计的高速铁路曲线控制点识别方法[J]. 铁道标准设计, 2026,70(4):40-47. DOI： 10.13238/j.issn.1004-2954.202406190007.

XIE Zhifeng, WEI Hui, ZHU Hongtao, et al. High-Speed Railway Curve Control Point Identification Method Based on PELT and Robust Estimation[J]. Railway Standard Design, 2026, 70(4): 40-47. DOI： 10.13238/j.issn.1004-2954.202406190007.

摘要

为能在天窗作业条件下准确掌握高速铁路曲线状态，并为之后的维修作业提供依据，利用部分轨道内部几何参数与曲线控制点存在映射关系，设计了一种基于相对测量数据的高速铁路曲线控制点识别方法。首先，计算并比较不同类型轨道内部几何参数的信噪比，优选出线形特征较明显的水平（超高）参数作为识别方法的输入数据；其次，依据高速铁路曲线线形类型及输入数据的特征，设置PELT搜索算法的目标函数及识别策略，实现水平（超高）数据的概略分割与曲线控制点初次识别；然后，对概略分割的水平（超高）数据进行分段线性稳健估计，实现变点里程的精确估计与曲线控制点二次识别，以减小异常数据对识别精度影响；最后，基于模拟和某试验线的水平（超高）数据，验证了识别方法在蒙特卡洛仿真实验中的正确性及实测应用中的有效性。仿真实验结果显示，受检信号信噪比不小于40.09 dB时，控制点识别总误差估计为0.26 m；实测实验结果显示，往返数据识别误差≤1.346 m，而单程数据识别误差一般不大于0.2 m，识别结果具有较高的重复性。由于识别数据来源于轨道内部几何参数，避免了外部几何参数的测量，因此具有较高的效率，可用于高铁日常轨道质量评价并为轨道维修作业提供依据。

Abstract

In order to accurately grasp the curve status of a high-speed railway under maintenance window conditions and provide a basis for the subsequent maintenance operations，a method of identifying the curve control points of high-speed railway based on relative measurement data was designed by utilizing the mapping relationship between some of the internal geometric parameters of the railway track and curve control points. Firstly，the signal-to-noise ratio（SNR）of the internal geometric parameters of different types of the railway track was calculated and compared，and the horizontal（superelevation）parameters with more obvious alignment characteristics were selected as the input data of the identification method. Secondly，according to the alignment type of the high-speed railway curve and the characteristics of the input data，the objective function and identification strategy of the PELT search algorithm were set to realize the rough segmentation of the horizontal（superelevation）data and the initial identification of curve control points. Then，the segmented horizontal（superelevation）data were subjected to piecewise linear robust estimation to achieve accurate estimation of change-point mileage and secondary identification of curve control points，so as to reduce the influence of abnormal data on identification accuracy. Finally，based on the simulation and the horizontal（superelevation）data of a test line，the correctness of the identification method in the Monte Carlo simulation experiment and its effectiveness in field application were verified. The results of simulation experiments showed that when the SNR of the detected signal was not less than 40.09 dB，the total error of the control point identification was estimated to be 0.26 m. The results of field experiments showed that the identification error of round-trip data was no greater than 1.346 m，while the identification error of one-way data was generally not more than 0.2 m，and the identification results showed high repeatability. Since the identification data are derived from the internal geometric parameters of the railway track，the method demonstrates efficiency and can be used in the daily track quality evaluation of high-speed railways as well as providing a basis for track maintenance operations.

关键词

Keywords

references

国家铁路局.高速铁路设计规范：TB 10621—2014[S].北京：中国铁道出版社，2015.

王开云，吕凯凯.高速铁路线路空间线形的动力学评价指标体系研究[J].西南交通大学学报，2016，51（2）:227-235.

WANG Kaiyun，LYU Kaikai. Dynamic Evaluation Index System for Spatial Alignment of High-speed Railway[J]. Journal of Southwest Jiaotong University，2016，51（2）:227-235.

魏庆朝，李鸣，吕希奎，等.高速铁路三维线形参数及其对列车运动的影响[J].铁道工程学报，2019，36（1）:31-37.

WEI Qingchao，LI Ming，LV Xikui，et al. Three-dimensional Alignment Parameters of the High-speed Railway and Their Influence on Motion of Train[J]. Journal of Railway Engineering Society，2019，36（1）:31-37.

时瑾，孙征南，孙宪夫，等.现行350 km/h高铁线路技术条件运营400 km/h高速列车适应性研究[J].铁道科学与工程学报，2020，17（9）:2171-2180.

SHI Jin，SUN Zhengnan，SUN Xianfu，et al. Research on the Adaptability of High-speed Train Running at 400 km/h Under Current 350km/h High-speed Railway Technical Conditions[J]. Journal of Railway Science and Engineering，2020，17（9）:2171-2180.

LONG X Y，WEI Q C，ZHENG F Y. Dynamic Analysis of Railway Transition Curves[J]. Proceedings of the Institution of Mechanical Engineers，Part F: Journal of Rail and Rapid Transit，2010，224（1）:1-14.

王平，杨荣山.轨道工程[M].北京：机械工业出版社，2021:391-401.

中国国家铁路集团有限公司.高速铁路线路维修规则：TG/GW115—2023[S].北京：中国铁道出版社，2023.

LIU J，CAI B G，WANG J. Electronic Track Map Building for Satellite-based High Integrity Railway Train Positioning[J]. International Journal on Smart Sensing and Intelligent Systems，2013，6（2）:610-629.

GARACH L，DE OÑA J，PASADAS M. Mathematical Formulation and Preliminary Testing of a Spline Approximation Algorithm for the Extraction of Road Alignments [J ]. Automation in Construction，2014，47:1-9.

高鸿剑，缪鹍.基于小波去噪的既有铁路平面重构方法[J].铁道科学与工程学报，2022，19（12）:3600-3611.

GAO Hongjian，MIAO Kun. A Method for Recreating Horizontal Alignment of Existing Railway Based on Wavelet Denoising[J]. Journal of Railway Science and Engineering，2022，19（12）:3600-3611.

缪鹍，田家凯，杨小礼.基于PSO的既有线曲线整正方法[J].中国铁道科学，2014，35（3）:8-14.

MIAO Kun，TIAN Jiakai，YANG Xiaoli. Realignment Method for Existing Railway Curve Based on PSO [J]. China Railway Science，2014，35（3）:8-14.

马洪磊，刘成龙.基于选权迭代法的既有铁路平面线形拟合方法[J].铁道科学与工程学报，2016，13（12）:2375-2380.

MA Honglei，LIU Chenglong. The Linear Fitting Method of Existing Railway Horizontal Align Ment Based on Iteration Method with Variable Weights[J]. Journal of Railway Science and Engineering，2016，13（12）:2375-2380.

LI W，PU H，SCHONFELD P，et al. A Method for Automatically Recreating the Horizontal Alignment Geometry of Existing Railways [J]. Computer-Aided Civil and Infrastructure Engineering，2019，34（1）:71-94.

陶维杰，蔡伯根，王剑，等.数字轨道地图平面线形特征提取方法研究[J].铁道学报，2019，41（9）:81-87.

TAO Weijie，CAI Baigen，WANG Jian，et al. Research on Geometry Extraction of Railway Horizontal Curves for Digital Track Map[J]. Journal of the China Railway Society，2019，41（9）:81-87.

SONG Z F，FANG T，SCHONFELD P，et al. Effect of Point Configurations on Parameter Estimation Analysis of Circles[J]. Journal of Surveying Engineering，2021，147（3）:04021010.

SONG T R，SCHONFELD P，PU H. A Review of Alignment Optimization Research for Roads，Railways and Rail Transit Lines [J]. IEEE Transactions on Intelligent Transportation Systems，2023，24（5）:4738-4757.

PU H，FU H D，SONG T R，et al. Recreating Horizontal Alignments of Existing Railways with a Hybrid Analytic and Harmony Search Algorithm [J]. Engineering Applications of Artificial Intelligence，2024，128:107354.

李颖，程朝阳，陆永彪，等.轨道曲线参数检测算法优化[J].铁道建筑，2023，63（10）:1-5.

LI Ying，CHENG Zhaoyang，LU Yongbiao，et al. Optimization of Algorithm for Detecting Track Curve Parameters[J]. Railway Engineering，2023，63（10）:1-5.

程朝阳，王昊，李颖，等.纵断面曲率检测及线形识别方法[J].铁道建筑，2023，63（8）:40-43.

CHENG Zhaoyang，WANG Hao，LI Ying，et al. Longitudinal Section Curvature Detection and Profile Recognition Method[J]. Railway Engineering，2023，63（8）:40-43.

殷华，朱洪涛，王志勇.基于奇异滤波及稳健回归的轨道曲线主点定位方法研究[J].铁道标准设计，2017，61（2）:43-47.

YIN Hua，ZHU Hongtao，WANG Zhiyong. Localization Algorithm for Track Curve Feature Point Based on Singular Value Decomposition and Robust Regression[J]. Railway Standard Design，2017，61（2）:43-47.

熊丽娟，朱洪涛，魏晖，等.基于轨道相对测量数据识别曲线关键参数的研究[J].铁道学报，2018，40（4）:127-134.

XIONG Lijuan，ZHU Hongtao，WEI Hui，et al. Research on Curve Key-parameter Identification Based Upon Track Relative Survey Data [J]. Journal of the China Railway Society，2018，40（4）:127-134.

易思蓉.铁路选线设计[M]. 4版.成都：西南交通大学出版社，2017:96-101.

魏晖，朱洪涛，赵国堂，等.基于中点弦测模型的无砟轨道精调量迭代求解[J].西南交通大学学报，2015，50（1）:131-136.

WEI Hui，ZHU Hongtao，ZHAO Guotang，et al. Iterative Algorithm of HSR Ballastless Track Realignment Calculation Based on MCO Model[J]. Journal of Southwest Jiaotong University，2015，50（1）:131-136.

GAO Z，XIAO X，FANG Y P，et al. A Selective Review on Information Criteria in Multiple Change Point Detection[J]. Entropy，2024，26（1）:50.

HAYNES K，FEARNHEAD P，ECKLEY I A. A Computationally Efficient Nonparametric Approach for Changepoint Detection[J]. Statistics and Computing，2017，27（5）:1293-1305.

TRUONG C，OUDRE L，VAYATIS N. Selective Review of Offline Change Point Detection Methods [J ]. Signal Processing，2020，167:107299.

茆诗松，王静龙，濮晓龙.高等数理统计[M]. 2版.北京：高等教育出版社，2006:147-158.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

高速铁路高架线地面振动测试评价及特征分析

基于雷视融合的高铁周界人员入侵检测方法

高铁列车动载作用下膨胀土隧道动力响应及服役性能

高速铁路多塔矮塔斜拉桥设计参数影响与敏感性分析

福厦高铁安海湾大跨斜拉特大桥无砟轨道施工竖向线形控制研究