| 2,075 | 36 | 629 |
| 下载次数 | 被引频次 | 阅读次数 |
随着铁路信息化进程的加快,铁路大数据与人工智能技术有力地支撑了我国铁路的高质量发展。本文回顾了近10年机器学习方法在铁路列车调度调整领域的主要研究成果,将列车调度调整问题分为列车晚点状态分析与评估、列车晚点传播预测和列车运行调整智能化决策三个方面,分别总结和分析了机器学习方法在上述三方面的应用情况。在列车晚点状态分析与评估方面,既有研究主要集中于传统统计分析,其描述和预测性能往往有限。在列车晚点传播预测方面,传统机器学习和深度学习方法被应用于晚点致因、晚点持续时长、列车晚点状态演化及晚点恢复预测问题的建模。在列车运行调整智能化决策方面,既有研究主要侧重于运用强化学习、模糊神经网络方法建模,机器学习方法应用于列车调度辅助决策系统仍是研究的主要难点和关键。在归纳了既有研究特征的基础上,展望了机器学习方法在铁路调度调整研究方面的最新动向。以深度学习为代表的高级机器学习应用于列车调度调整智能决策将是未来的发展重点。
Abstract：With the rapid development of railway informatization, big data and artificial intelligence technology are providing strong support for the high-quality development of China's railway system. The major studies of the application of machine learning methods to railway train operation adjustment in the past 10 years are reviewed. The problem of train operation adjustment is divided into three topics: analysis and evaluation of the train delay state, prediction of train delay propagation,and train operation adjustment and intelligent decision-making. The application of machine learning methods in these three topics is summarized and analyzed. Existing research on train delay analysis and evaluation mainly focuses on traditional statistical analysis, and its description and prediction performance are often limited. Traditional machine learning and deep learning methods have been applied to the modeling of the causes and duration of train delay, the evolution of train delays, and the prediction of train delay recovery. The existing research on the intelligent decision-making for train operation adjustment mainly focuses on applying reinforcement learning and fuzzy neural networks to establish models, while the application of machine learning in train dispatching support systems development remains the main difficulty and the key problem. The characteristics of the existing research are summarized. On this basis, this paper looks forward to the latest trend of using machine learning methods in the research of railway train operation adjustment. The application of advanced machine learning methods represented by deep learning in intelligent decision-making for train operation adjustment will be highly focused in the future.
[1]“智能高铁战略研究（2035）”通过工程院结题评审[EB/OL].[2020-01-15].http://www.chinarevit.com/revit-55316-1-1.html.
[2] GRAVES A, MOHAMED A R, HINTON G. Speech recognition with deep recurrent neural networks[C]//2013IEEE International Conference on Acoustics, Speech and Signal Processing. Vancouver, Canada:IEEE, 2013:6645-6649.
[3] SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[EB/OL].(2015-04-10)[2021-03-15] https://arxiv. org/abs/1409.1556.
[4]赵雁飞.行车指挥自动化几个关键问题的研究[D].北京：北方交通大学, 2000.
[5]中国科学院. 2019全球人工智能发展白皮书（2019）[R].北京：中国科学院, 2019.
[6]赛迪顾问股份有限公司. 2018年中国大数据产业生态白皮书[R].北京：赛迪顾问股份有限公司，2018.
[7] KECMAN P, GOVERDE R. Process mining approach for recovery of realized train paths and route conflict identification[C]//Transportation Research Board TRB:92nd Annul meeting. Waslaington, USA:TRB, 2013.
[8]廖志林,唐伟忠.铁路调度综合管理平台应用研究[J].铁路计算机应用, 2018, 27(4):25-28.
[9] KIM K, SEOL S, KONG S H. High-speed train navigation system based on multi-sensor data fusion and map matching algorithm[J]. International Journal of Control,Automation and Systems, 2015, 13(3):503-512.
[10]雷明.机器学习——原理、算法与应用[M].北京:清华大学出版社, 2019.
[11] GRAFFAGNINO T. Ensuring timetable stability with train traffic data[J]. Computers in Railways XIII:Computer System Design and Operation in the Railway and Other Transit Systems, 2013,127:427-438.
[12]刘岩,郭竞文,罗常津,等.列车运行实绩大数据分析及应用前景展望[J].中国铁路, 2015(6):70-73.
[13] WEN C, LI Z, LESSAN J, et al. Statistical investigation on train primary delay based on real records:evidence from Wuhan–Guangzhou HSR[J]. International Journal of Rail Transportation, 2017, 5(3):170-189.
[14]徐传玲,文超,胡瑞,等.高速铁路列车连带晚点产生机理及其判定[J].交通运输工程与信息学报, 2020, 18(4):31-37.
[15]庄河,文超,李忠灿,等.基于高速列车运行实绩的致因-初始晚点时长分布模型[J].铁道学报, 2017, 39(9):25-31.
[16]文超,李忠灿,黄平,等.高速铁路初始晚点致因-影响列车数分布模型[J].西南交通大学学报, 2018, 53(6):1261-1269.
[17] MARKOVI?N, MILINKOVI?S, TIKHONOV K S, et al. Analyzing passenger train arrival delays with support vector regression[J]. Transportation Research Part C:Emerging Technologies, 2015, 56:251-262.
[18] MURALI P, DESSOUKY M, ORDó?EZ F, et al. A delay estimation technique for single and double-track railroads[J]. Transportation Research Part E:Logistics and Transportation Review, 2010, 46(4):483-495.
[19] LEE W H, YEN L H, CHOU C M. A delay root cause discovery and timetable adjustment model for enhancing the punctuality of railway services[J]. Transportation Research Part C:Emerging Technologies, 2016, 73:49-64.
[20] CERRETO F, NIELSEN B F, NIELSEN O A, et al. Application of data clustering to railway delay pattern recognition[J/OL]. Journal of Advanced Transportation,2018:18(2018-04-29). https://doi. org/10.1155/2018/6164534.
[21]张琦,陈峰,张涛,等.高速铁路列车连带晚点的智能预测及特征识别[J].自动化学报, 2019, 45(12):2251-2259.
[22] CHEBOTAREV V, DAVYDOV B, KABLUKOVA K.Probabilistic model of delay propagation along the train flow[J]. Probabilistic Modeling in System Engineering,IntechOpen, 2018:171-193.
[23] HARROD S, CERRETO F, NIELSEN O A. A closed form railway line delay propagation model[J]. Transportation Research Part C:Emerging Technologies, 2019,102:189-209.
[24] HARROD S, POURNARAS G, NIELSEN B F. Distribution fitting for very large railway delay data sets with discrete values[J]. Danish Journal of Transportation Research–Dansk Tidsskrift for Transportforskning, 2019,1:1-12.
[25] HUANG P, WEN C, PENG QY, et al. Modeling the Influence of Disturbances in High-Speed Railway Systems[J/OL]. Journal of Advanced Transportation, 2019:13.(2019-05-19).https://doi.org/10.1155/2019/8639589.
[26] ONETO L, BUSELLI I, SANETTI P, et al. Restoration time prediction in large scale railway networks:big data and Interpretability[C]//Recent Advances in Big Data and Deep Learning:Proceedings of the INNS Big Data and Deep Learning Conference INNSBDDL2019. Genova, Italy:Springer, 2019, 1:136-141.
[27] ZILKO A, HANEA A, KUROWICKA D, et al. Nonparametric Bayesian network to forecast railway disruption lengths[C]//PROC. 2nd International Conference on Railway Technology:Research, Development and Maintenance(Railways 2014). Ajaccio, France:PROC, 2014.
[28] ZILKO A A, KUROWICKA D, GOVERDE R M. Modeling railway disruption lengths with copula Bayesian networks[J]. Transportation Research Part C:Emerging Technologies, 2016, 68:350-368.
[29]汤轶雄,徐传玲,文超,等.高铁故障晚点时间预测的支持向量回归模型[J].中国安全科学学报, 2019, 29(S2):18-23.
[30] WANG R, WORK D B. Data driven approaches for passenger train delay estimation[C]//IEEE. 2015 IEEE 18th International Conference on Intelligent Transportation Systems. Gran Canaria, Spain:IEEE, 2015:535-540.
[31] LI D, DAAMEN W, GOVERDE R M. Estimation of train dwell time at short stops based on track occupation event data:a study at a Dutch railway station[J]. Journal of Advanced Transportation, 2016, 50(5):877-896.
[32] GUO J, MENG L, KECMAN P, et al. Modeling delay relations based on historical train monitoring data:a Chinese high-speed railway case[C]//6th International Conference on Railway Operations Modelling and Analysis.Tokyo, Japan:ROMA, 2015.
[33] KECMAN P, GOVERDE R M. Predictive modelling of running and dwell times in railway traffic[J]. Public Transport, 2015, 7(3):295-319.
[34] PONGNUMKUL S, PECHPRASARN T, KUNASETH N, et al. Improving arrival time prediction of Thailand’s passenger trains using historical travel times[C]//11th International Joint Conference on Computer Science and Software Engineering. Chonburi, Thailand:JCSSE, 2014:307-312.
[35] KECMAN P, CORMAN F, MENG L. Train delay evolution as a stochastic process[C]//6th International Conference on Railway Operations Modelling and Analysis Rail. Tokyo, Japan:ROMA, 2015.
[36] BARTA J, RIZZOLI A E, SALANI M, et al. Statistical modelling of delays in a rail freight transportation network[C]//Proceedings of the 2012 Winter Simulation Conference(WSC). Berlin, Germany:WSC, 2012:1-12.
[37] GAURAV R, SRIVASTAVA B. Estimating train delays in a large rail network using a zero shot markov model[C]//2018 21st International Conference on Intelligent Transportation Systems ITSC. Maui, USA:ITSC, 2018:1221-1226.
[38] CORMAN F, KECMAN P. Stochastic prediction of train delays in real-time using Bayesian networks[J]. Transportation Research Part C:Emerging Technologies,2018, 95:599-615.
[39] ONETO L, FUMEO E, CLERICO G, et al. Advanced analytics for train delay prediction systems by including exogenous weather data[C]//2016 IEEE International Conference on Data Science and Advanced Analytics.Montreal, Canada:DSAA, 2016:458-467.
[40] ARSHAD M, AHMED M. Train delay estimation in indian railways by including weather factors through machine learning techniques[J]. Recent Advances in Computer Science and Communications, 2021, 14(4):1300-1307.
[41] YAGHINI M, KHOSHRAFTAR M M, SEYEDABADI M. Railway passenger train delay prediction via neural network model[J]. Journal of Advanced Transportation,2013, 47(3):355-368.
[42] ONETO L, FUMEO E, CLERICO G, et al. Delay prediction system for large-scale railway networks based on big data analytics[C]//INNS Conference on Big Data.Thessaloniki, Greece:INNS, 2016:139-150.
[43] ONETO L, FUMEO E, CLERICO G, et al. Dynamic delay predictions for large-scale railway networks:deep and shallow extreme learning machines tuned via thresholdout[J]. IEEE Transactions on Systems, Man, and Cybernetics:Systems, 2017, 47(10):2754-2767.
[44]孙略添,宋瑞,何世伟,等.技术站列车晚点时间预测方法[J].北京交通大学学报, 2018, 42(1):94.
[45] ONETO L, FUMEO E, CLERICO G, et al. Train delay prediction systems:a big data analytics perspective[J].Big Data Research, 2018, 11:54-64.
[46] WEN C, MOU W, HUANG P, et al. A predictive model of train delays on a railway line[J]. Journal of Forecasting, 2020, 39(3):470-488.
[47] HUANG P, WEN C, FU L, et al. A deep learning approach for multi-attribute data:a study of train delay prediction in railway systems[J]. Information Sciences,2020, 516:234-253.
[48] LI Z, HUANG P, WEN C, et al. Predictive models for influence of primary delays using high-speed train operation records[J]. Journal of Forecasting, 2020, 39(8):1198-1212.
[49] KHADILKAR H. Data-enabled stochastic modeling for evaluating schedule robustness of railway networks[J].Transportation Science, 2016, 51(4):1161-1176.
[50] YANG X, HOU Y, LI L. Buffer time allocation according to train delay expectation at stations[J]. International Journal of Rail Transportation, 2020, 8(3):249-263.
[51] JIANG C, HUANG P, LESSAN J, et al. Forecasting primary delay recovery of high-speed railway using multiple linear regression, supporting vector machine, artificial neural network, and random forest regression[J]. Canadian Journal of Civil Engineering, 2019, 46(5):353-363.
[52]?AHIN I. Markov chain model for delay distribution in train schedules:assessing the effectiveness of time allowances[J]. Journal of Rail Transport Planning&Management, 2017, 7(3):101-113.
[53] HUANG P, WEN C, PENG Q Y, et al. A data-driven time supplements allocation model for train operations on high-speed railways[J]. International Journal of Rail Transportation, 2019, 7(2):140-157.
[54] MARTIN L J. Predictive reasoning and machine learning for the enhancement of reliability in railway systems[C]//International Conference on Reliability, Safety, and Security of Railway Systems. Paris, France. 2016:178-188.
[55]胡雨欣,彭其渊,鲁工圆,等.基于初始晚点和冗余时间的列车晚点恢复时间预测模型[J].交通运输工程与信息学报, 2020, 18(2):93-102.
[56] CACCHIANI V, HUISMAN D, KIDD M, et al. An overview of recovery models and algorithms for real-time railway rescheduling[J]. Transportation Research Part B:Methodological, 2014, 63:15-37.
[57] CADARSO L, MARíNá. Recovery of disruptions in rapid Transit networks with origin-destination demand[J]. Procedia-Social and Behavioral Sciences, 2014, 111:528-537.
[58]占曙光,赵军,彭其渊,等.高速铁路区间能力全失效条件下列车运行实时调整研究[J].铁道学报, 2015, 37(11):1-9.
[59] KRASEMANN J T. Design of an effective algorithm for fast response to the re-scheduling of railway traffic during disturbances[J]. Transportation Research Part C:Emerging Technologies, 2012, 20(1):62-78.
[60]王涛,张琦,赵宏涛,等.基于替代图的列车运行调整计划编制及优化方法[J].中国铁道科学, 2013, 34(5):126-133.
[61]庄河,何世伟,戴杨铖.高速铁路列车运行调整的模型及其策略优化方法[J].中国铁道科学, 2017, 38(2):118-126.
[62] DüNDAR S,?AHIN I. Train re-scheduling with genetic algorithms and artificial neural networks for single-track railways[J]. Transportation Research Part C:Emerging Technologies, 2013, 27:1-15.
[63] ONETO L, BUSELLI I, LULLI A, et al. Train overtaking prediction in railway networks:a big data perspective[C]//INNS Big Data and Deep Learning conference.Genova, Italy:Springer, 2019:142-151.
[64]?EMROV D, MARSETI?R,?URA M, et al. Reinforcement learning approach for train rescheduling on a single-track railway[J]. Transportation Research Part B:Methodological, 2016, 86:250-267.
[65] KHADILKAR H. A scalable reinforcement learning algorithm for scheduling railway lines[J]. IEEE Transactions on Intelligent Transportation Systems, 2018, 20(2):727-736.
[66] SUN J, WANG S. Intelligent train dispatching for highspeed railway based on fuzzy neural network[C]//20102nd International Conference on Advanced Computer Control. Shenyang:CACC, 2010:222-225.
[67] NARAYANASWAMI S, RANGARAJ N. A MAS architecture for dynamic, realtime rescheduling and learning applied to railway transportation[J]. Expert Systems with Applications, 2015, 42(5):2638-2656.
[68] ZHU T, MERA J M, SUAREZ B, et al. An agent-based support system for railway station dispatching[J]. Expert Systems with Applications, 2016, 61:39-52.
[69] ZHU T, DE PEDRO J M M S. Railway traffic conflict detection via a state transition prediction approach[J].IEEE Transactions on Intelligent Transportation Systems, 2017, 18(5):1268-1278.
[70]唐涛,李开成,宿帅,等.瑞士铁路调度指挥系统特点分析及对我国启示[J].中国铁路, 2019(11):18-23.
[71] GHOFRANI F, HE Q, GOVERDE R M, et al. Recent applications of big data analytics in railway transportation systems:a survey[J]. Transportation Research Part C:Emerging Technologies, 2018, 90:226-246.
[72]阿里研究院.从工具革命到决策革命——通向智能制造的转型之路[R].杭州：阿里研究院，2019.
[73]李平,邵赛,薛蕊,等.国外铁路数字化与智能化发展趋势研究[J].中国铁路, 2019(02):25-31.
[74]阿里巴巴达摩院.达摩院2020十大科技趋势[R].杭州：阿里巴巴达摩院，2020.
基本信息:
DOI：10.19961/j.cnki.1672-4747.2021.04.021
中图分类号:TP181;U292.4
引用信息:
[1]文超,李津,李忠灿,等.机器学习在铁路列车调度调整中的应用综述[J],2022,20(01):1-14.DOI:10.19961/j.cnki.1672-4747.2021.04.021.
基金信息:
国家自然科学基金项目（71871188,U1834209）;; 中国国家铁路集团有限公司科技研究计划项目（P2020X016）;; 山东省自然科学基金项目（ZR2019PG008）