Adaptive modeling method based on the Fast-MCD to analyze railway track irregularity deterioration
YANG Yaqin1, XU Peng1, WU Xishui2
1. School of Traffic and Transportation, Beijing Jiaotong University, Beijing 100044, China; 2. Infrastructure Management Department, China National Railway Group Co., Ltd., Beijing 100844, China
Abstract：Railway track predictive maintenance needs to identify the track irregularity deterioration mode. For this, the track irregularity deterioration process is split into maintenance periods according to the maintenance records to analyze the deterioration mode. A rail track irregularity deterioration adaptive piecewise modeling framework was developed. In order to improve the efficiency and accuracy of the framework, a two-level fast minimum covariance determinant (Fast-MCD) was developed based on the Mahalanobis distance. The critical values for each level were determined for a specific case by measurements along the Nanchang-Fuzhou railway from 2013 to 2020. Analysis of the data showed that the efficiency is improved by nearly 50% and the identification accuracy of the maintenance dates is improved by nearly 45.3% through use of the two-level Fast-MCD. The track irregularity deterioration mode was further studied by using this framework to analyze the track longitudinal deterioration of each track unit for each maintenance period with a total of nearly 6 000 maintenance periods. The goodness-of-fit and prediction capabilities of linear, exponential, and logarithmic functions were compared for each maintenance period. The results show that the linear function best describes the track longitudinal irregularity deterioration.
[1] QUIROGA L M, SCHNIEDER E. A heuristic approach to railway track maintenance scheduling[M]//NING B, BREBBIA C A. Computers in Railways XII. Boston:WIT Press, 2010:687-699. [2] VEIT P, MARSCHNIG S. Sustainability in track:A precondition for high speed traffic[C]//Proceedings of 2010 Joint Rail Conference. Urbana, USA:ASME, 2010. [3] 郭然, 韩宝明, 李得伟, 等. 具有更新机制的铁路轨道不平顺灰色预测模型[J]. 中南大学学报(自然科学版), 2013, 44(10):4334-4341. GUO R, HAN B M, LI D W, et al. Grey prediction model for track irregularity with update mechanism[J]. Journal of Central South University (Science and Technology), 2013, 44(10):4334-4341. (in Chinese) [4] ANDRADE A R, TEIXEIRA P F. Biobjective optimization model for maintenance and renewal decisions related to rail track geometry[J]. Transportation Research Record:Journal of the Transportation Research Board, 2011, 2261(1):163-170. [5] 许玉德, 吴纪才. 利用线性预测模型分析轨道不平顺发展[J]. 石家庄铁道学院学报, 2005, 18(1):6-9. XU Y D, WU J C. Analysis on development of track irregularities with linear forecast model[J]. Journal of Shijiazhuang Railway Institute, 2005, 18(1):6-9. (in Chinese) [6] 杨飞. 基于高低不平顺的线路捣固作业维修标准及决策技术研究[J]. 铁道建筑, 2017(7):131-135. YANG F. Research on maintenance standard and decision-making technique for tamping operation based on longitudinal irregularity[J]. Railway Engineering, 2017(7):131-135. (in Chinese) [7] SOLEIMANMEIGOUNI I, XIAO X, AHMADI A, et al. Modelling the evolution of ballasted railway track geometry by a two-level piecewise model[J]. Structure and Infrastructure Engineering, 2018, 14(1):33-45. [8] DROŹDZIEL J, SOWIŃSKI B. Simulation of railway track deterioration influenced by ballast stiffness and dry friction[M]//ALLAN J, ARIAS E, BREBBIA C A, et al. Computers in Railways XI. Boston:WIT Press, 2008:693-702. [9] 杨雅琴, 徐鹏, 李晔, 等. 适用复杂劣化趋势的轨道不平顺鲁棒建模方法[J]. 交通运输系统工程与信息, 2020, 20(5):156-162. YANG Y Q, XU P, LI Y, et al. Robust modeling method for track irregularity of complicated deterioration trend[J]. Journal of Transportation Systems Engineering and Information Technology, 2020, 20(5):156-162. (in Chinese) [10] RISSANEN J. Stochastic complexity in statistical inquiry theory[M]. River Edge:World Scientific Publishing Co., 1989. [11] 中国国家铁路集团有限公司. 高速铁路线路维修规则(征求意见稿)[M]. 北京:中国铁道出版社, 2020. China National Railway Group Co., Ltd. Rules of high-speed railway track maintenance (draft for comments)[M]. Beijing:China Railway Publishing House, 2020. (in Chinese) [12] MAHALANOBIS P C. On the generalized distance in statistics[J]. Proceedings of the National Institute of Sciences of India, 1936(2):49-55. [13] ROUSSEEUW P J. Least median of squares regression[J]. Journal of the American Statistical Association, 1984, 79(388):871-880. [14] ROUSSEEUW P J, VAN DRIESSEN K. A fast algorithm for the minimum covariance determinant estimator[J]. Technometrics, 1999, 41(3):212-223. [15] XU P, SUN Q X, LIU R K, et al. Optimizing the alignment of inspection data from track geometry cars[J]. Computer-Aided Civil and Infrastructure Engineering, 2015, 30(1):19-35.
2022, Vol. 62 
