自公共交通优先发展战略确立以来，城市公共交通系统的信息化、智能化水平得到极大提升。其中，自动车辆定位（Automatic Vehicle Location，简称AVL）技术的逐渐成熟以及智能车载终端在公交车辆的普及应用，为公交行程行程时间研究提供了良好的基础数据环境。本研究从历史公交运行轨迹的数据挖掘出发，深入探讨公交运行过程中路径行程时间存在的机理特征，即公交路径行程时间比例特性，并以其为理论基础搭建适用于不同数据环境下的公交到站时间预测模型。

首先，从理论研究层面提出并证明了公交路径行程时间比例特性。从分析公交运行过程入手，结合公交实际运行情况及相关数学定理，将站间行驶时间以及站点停靠时间进行模型化，进而推导出公交班次间路径行程时间的比值表达式，从理论层面提出并证明了公交路径行程时间比例特性的存在。接着，考虑到理论证明中存在部分理想性假设，为了进一步验证比例特性存在的真实性，利用江阴市K19路，上海市210路两条线路的历史公交运行班次数据集，选取三种典型参照班次，包括前一班次、全样本计划班次、小时样本计划班次，进行实证验证。结果发现：1）无论是选择前一班次，抑或是基于全样本计划班次，还是基于小时样本计划班次作为参照班次，班次路径行程时间之间均呈现出显著的比例特性，即随着公交车辆行驶，其与参照班次路径行程时间之间的比例系数逐渐收敛，并趋于稳定；2）比较而言，选取两种计划班次作为参照班次时，路径行程时间比例系数的波动更小，收敛速度更快。

然后，从应用层面提出完整数据环境下基于比例特性的到站时间预测方法。以公交路径行程时间比例特性为理论基础，提出一种可移植性与可解释性强的公交到站时间预测模型，即比例模型，并将其在完整数据环境下的预测性能与现有模型进行评估对比，结果显示：1）比例模型的平均绝对误差（Mean Absolute Error，简称MAE）呈现出沿线迅速减小的趋势，预测结果的准确性越来越高，而平均绝对百分比误差（Mean Absolute Percentage Error，简称MAPE）则在线路中间区域时最小，预测结果的精确性最高。其中线路K19中大部分站点的MAPE值低于10%，线路210中大部分站点的MAPE值低于20%。2）比较而言，以小时样本计划班次为参照班次的比例模型预测性能略优于以全样本计划班次为参照班次的比例模型，但并不显著；两种比例模型在大部分站点的预测性能均接近于现有模型，但其在始发站点附近区域劣势较为显著。基于比例系数沿线收敛特征，提出比例系数边界约束优化方法，显著提升了比例模型在始发站点附近区域的预测性能。

最后，从应用层面对稀疏数据环境下比例模型的预测性能进行评价。考虑到实际运营过程中受诸多内外部环境影响，公交轨迹数据的完整性往往难以保证，由此本研究对完整数据环境下的比例模型进行改进，使其同样适用于在稀疏数据环境下进行公交到站时间预测，并对其在数据丢失环境下的预测性能进行了评估，结果显示：在不同数据丢失率下，比例模型的预测性能退化程度并不显著，即其在稀疏数据环境下具有较强的适应性，具有极强的实用性。

[1]国务院办公厅. 关于优先发展公共交通意见的通知. (国办发〔2005〕46 号), 2005.

[2]Rahman M M, Wirasinghe S C, Kattan L. Analysis of bus travel time distributions for varying horizons and real-time applications[J]. Transportation Research Part C: Emerging Technologies, 2018, 86: 453-466.

[3]Jiaman M, Jeffrey C, Goce R, et al. Bus travel time prediction with real-time traffic information[J]. Transportation Research Part C: Emerging Technologies, 2019, 105: 536-549.

[4]Faruk S, Yigit A, Adnan K. Hybrid time series forecasting methods for travel time prediction[J]. Physica A: Statistical Mechanics and its Applications, 2021, 579: 126-134.

[5]刘明鑫. 基于AVL数据的城市道路公交行程时间预测方法[D]. 吉林大学, 2020.

[6]Yetiskul E, Senbil M. Public bus transit travel-time variability in Ankara (Turkey)[J]. Transport Policy, 2012, 23: 50-59.

[7]Zhang X, Ren Y, Huang B, et al. Analysis of time-varying characteristics of bus weighted complex network in Qingdao based on boarding passenger volume[J]. Physica A: Statistical Mechanics and its Applications, 2018, 506: 376-394.

[8]王芳杰, 王福建, 王雨晨, 等. 基于LightGBM算法的公交行程时间预测[J]. 交通运输系统工程与信息, 2019, 19(02): 116-121.

[9]Celan M, Lep M. Bus arrival time prediction using bus network data model and time periods[J]. Future Generation Computer Systems-the International Journal of Escience, 2020, 110: 364-371.

[10]Tirachini A. Estimation of travel time and the benefits of upgrading the fare payment technology in urban bus services[J]. Transportation Research Part C: Emerging Technologies, 2013, 30(5): 239-256.

[11]He P, Jiang G, Lam S K, et al. Learning heterogeneous traffic patterns for travel time prediction of bus journeys[J]. Information Sciences, 2019, 512(4): 1394-1406.

[12]Yu Z, Wood J S, Gayah V V, et al. Using survival models to estimate bus travel times and associated uncertainties[J]. Transportation Research Part C: Emerging Technologies, 2017, 74: 366-382.

[13]郭宪. 基于梯度提升决策回归树的公交行程时间预测方法[D]. 华中科技大学, 2019.

[14]宋现敏, 刘明鑫, 马林, 等. 基于极限学习机的公交行程时间预测方法[J]. 交通运输系统工程与信息, 2018, 18(5): 136-142+150.

[15]彭新建, 翁小雄. 基于萤火虫算法优化BP神经网络的公交行程时间预测[J].广西师范大学学报(自然科学版), 2017, 35(1): 28-36.

[16]Tang J, Yang Y, Qi Y. A hybrid algorithm for Urban transit schedule optimization[J]. Physica A: Statistical Mechanics and its Applications, 2018, 512: 745-755.

[17]Daganzo C F. A Headway-based Approach to Eliminate Bus Bunching: Systematic Analysis and Comparisons[J]. Transportation Research Part B: Methodological, 2009, 43(10): 913-921.

[18]Qiang M, Qu X. Bus dwell time estimation at bus bays: a probabilistic approach[J]. Transportation Research Part C: Emerging Technologies, 2013, 36(36): 61-71.

[19]肖梅, 王仲豪, 颜建强, 等. 基于西安城市公交IC卡数据的乘客上车耗时影响因素分析[J]. 重庆理工大学学报(自然科学), 2019, 33(01): 136-142.

[20]张梦洁. 基于生存分析的公交停靠耗费时间影响因素研究[D]. 西南交通大学, 2019.

[21]Gao Y, Chen L, Chen D, et al. How does a static measure influence passengers’ boarding behaviors and bus dwell time? Simulated evidence from Nanjing bus stops[J]. Transportation Research Part A: Policy and Practice, 2018, 110: 13-25.

[22]Jaiswal S, Bunker J, Ferreira L. The influence of platform walking on BRT station bus dwell time[J]. Journal of Transportation Engineering, 2010, 136(12): 1173-1179.

[23]姜景玲, 张亚南, 闫超. 中心城区公交站点停靠时间可靠性影响因素分析[J]. 公路交通科技, 2020, 37(S1): 136-141.

[24]Pan Y, Chen S, Li T, et al. Exploring spatial variation of the bus stop influence zone with multi-source data: A case study in Zhenjiang, China[J]. Journal of Transport Geography, 2019, 76: 166-177.

[25]Sun W, Schmocker J D, Fukuda K. Estimating the route-level passenger demand profile from bus dwell times[J]. Transportation Research Part C: Emerging Technologies, 2020, 130273.

[26]王旭, 陈旭梅, 寇伟彬, 等. 公交站点停靠时间预测模型研究[J]. 交通信息与安全, 2016, 34(02): 55-61.

[27]曹守华, 袁振洲, 赵丹. 城市轨道交通乘客上车时间特性分析及建模[J]. 铁道学报, 2009, 31(03): 89-93.

[28]Sun L, Tirachini A, Axhausen K W, et al. Models of bus boarding and alighting dynamics[J]. Transportation Research Part A: Policy and Practice, 2014, 69: 447-460.

[29]Milkovits M N. Modeling the Factors Affecting Bus Stop Dwell Time: Use of Automatic Passenger Counting, Automatic Fare Counting, and Automatic Vehicle Location Data[J]. Transportation Research Record, 2008, 2072(1): 125-130.

[30]Tirachini A. Bus dwell time: the effect of different fare collection systems, bus floor level and age of passengers[J]. Transportmetrica A-transport Science, 2013, 9(1-2): 28-49.

[31]Chen G, Chen W L, Zhang S Y, et al. Influence of Mobile Payment on Bus Boarding Service Time. Journal of Advanced Transportation, 2020: 1-12.

[32]Dueker K J, Kimpel T J, Strathman J G. Determinants of Bus Dwell Time[J]. Journal of Public Transportation, 2004, 7(1): 21-40.

[33]霍豪, 郑长江, 沈金星. 基于近邻和SVM的公交停靠时间预测[J]. 交通运输工程与信息学报, 2021, 19(3): 59-66.

[34]EI-Geneidy M A, Vijayakumar, et al. The effects of articulated buses on dwell and running times[J]. Journal of Public Transportation, 2011, 14(3): 63-86.

[35]吕伟, 李志红, 赵彩虹, 等. 城市公交车停靠时间特征的观测试验与实证研究[J]. 公路交通科技, 2019, 36(11): 90-96.

[36]汪磊, 左忠义. 基于 MLR 的公交车行程时间预测模型[J]. 大连交通大学学报, 2015, 36(2): 1-5.

[37]Tirachini A. Estimation of travel time and the benefits of upgrading the fare payment technology in urban bus services[J]. Transportation Research Part C: Emerging Technologies, 2013, 30(5): 239-256.

[38]Zhou Y Y, Yao L, Chen Y Y, et al. Bus arrival time calculation model based on smart card data[J]. Transportation Research Part C: Emerging Technologies, 2017: 81-96.

[39]Xu H T, Ying J. Bus arrival time prediction with real-time and historic data[J]. Journal of Networks Software Tools and Applications, 2017, 20(04): 3099-3106.

[40]Wei Y, Chen M. Forecasting the short-term metro passenger flow with empirical mode decomposition and neural networks[J]. Transportation Research Part C: Emerging Technologies, 2012, 21(1): 148-162.

[41]Mazloumi E, Rose G, Currie G, et al. Prediction intervals to account for uncertainties in neural network predictions: Methodology and application in bus travel time prediction[J]. Engineering Applications of Artificial Intelligence, 2011, 24(3): 534-542.

[42]Pan J, Dai X, Xu X, et al. A self-learning algorithm for predicting bus arrival time based on historical data model[C]. 2012 IEEE 2nd International Conference on Cloud Computing and Intelligence Systems. IEEE, 2012, 3: 1112-1116.

[43]Vanajakshi L, Rilett L. Support vector machine technique for the short-term prediction of travel time[C]. Intelligent Vehicles Symposium, 2007, 2007: 600-605.

[44]Yu B, Yang Z Z, Wang J. Bus travel-time prediction based on bus speed[J]. Proceedings of the ICE-Transport, 2010, 163(1): 3-7.

[45]Yang M, Chen C, Wang L, et al. Bus arrival time prediction using support vector machine with genetic algorithm. Neural Network World, 2016, 26(3), 205-217.

[46]于滨, 杨忠振, 林剑艺. 应用支持向量机预测公交车运行时间[J]. 系统工程理论与实践, 2007, 2007(4): 160-164+176.

[47]Sharmila R B, Velaga N R, Kumar A. SVM-based hybrid approach for corridor-level travel-time estimation[J]. IET Intelligent Transport Systems, 2019, 13(9): 1429-1439.

[48]李海姣, 陆建. 基于交通流理论的公交站点间行程时间预测[J]. 交通信息与安全, 2012, 30(3): 29-32.

[49]孙棣华, 赖云波, 廖孝勇, 等. 公交浮动车辆到站时间实时预测模型[J]. 交通运输工程学报, 2011, 11(2): 84-89.

[50]Chen G, Yang X, An J, et al. Bus-arrival-time prediction models: Link-based and section-based. Journal of transportation engineering, 2012, 138(1): 60-66.

[51]Kumar B A, Vanajakshi L, Subramanian S C. A hybrid model based method for bus travel time estimation[J]. Journal of Intelligent Transportation Systems, 2017, 1-17.

[52]Chung E H., Shalaby A. Expected time of arrival model for school bus transit using real-time global positioning system-based automatic vehicle location data[J]. Journal of Intelligent Transportation Systems, 2007, 11(4), 157-167.

[53]Cristóbal T, Padrón G, Quesada-Arencibia A, et al. Bus travel time prediction model based on profile similarity[J]. Sensors, 2019, 19(13): 2869.

[54]李文正. 使用ARIMA模型预测公交到站时间[D]. 清华大学, 2015.

[55]Suwardo, Napiah M, Kamaruddin I. Arima models for bus travel time prediction[J]. Institution of Engineers Malaysia, 2010, 71(2): 49-58.

[56]朱丽颖. 公交车辆行程时间预测方法研究[D]. 北京交通大学, 2010.

[57]周帅帅. 城市公交车辆到站时间预测方法研究[D]. 华东交通大学, 2017.

[58]宋爽. 基于SVM和Kalman滤波的公交到站时间预测方法研究[D]. 大连海事大学, 2018.

[59]Kumar S V, Chaitanya Dogiparthi K, Vanajakshi L, et al. Integration of exponential smoothing with state space formulation for bus travel time and arrival time prediction[J]. Transport, 2017, 32(4): 358-367.

[60]Han Q, Liu K, Zeng L, et al. A bus arrival time prediction method based on position calibration and LSTM. IEEE Access, 2020, 8: 42372-42383.

[61]Liu H G, Xu H Z, Yan Y, et al. Bus arrival time prediction based on LSTM and spatial-temporal feature vector[J]. IEEE ACCESS, 2020, 8: 11917-11929.

[62]Alam O, Kush A, Emami A, et al. Predicting irregularities in arrival times for transit buses with recurrent neural networks using GPS coordinates and weather data[J]. Journal of Ambient Intelligence and Humanized Computing, 2021, 12(7): 7813-7826.

[63]Wang D, Zhang J, Cao W, et al. When will you arrive? estimating travel time based on deep neural networks[C]. Thirty-Second AAAI Conference on Artificial Intelligence, 2018, 1-8.

[64]Petersen N C, Rodrigues F, Pereira F C. Multi-output bus travel time prediction with convolutional LSTM neural network[J]. Expert Systems with Applications, 2019, 120: 426-435.

[65]Chen X, Zhang S, Li L, et al. Adaptive rolling smoothing with heterogeneous data for traffic state estimation and prediction[J]. IEEE Transactions on Intelligent Transportation Systems, 2019, 20(4): 1247-1258.

[66]Cai X S. Collaborative prediction for bus arrival time based on GPS[J]. Journal of Central South University, 2014, 21(3): 1242-1248.

[67]郭敏, 蓝金辉, 李娟娟, 等. 基于灰色残差GM(1, N)模型的交通流数据恢复算法[J]. 交通运输系统工程与信息, 2012, 12(1): 42-47.

[68]王薇, 程泽阳, 刘梦依, 等. 基于时空相关性的交通流故障数据修复方法[J]. 浙江大学学报(工学版), 2017, 51(9): 1727-1734.

[69]Tak S, Woo S, Yeo H. Data-driven imputation method for traffic data in sectional units of road links[J]. IEEE Transactions on Intelligent Transportation Systems, 2016, 17(6): 1762-1771.

[70]郑业晴, 朱欣焰, 张发明, 等. 基于路网相似性的路段行程时间估计[J]. 计算机应用研究, 2018, 35(6): 1681-1685.

[71]Wang Y, Zhang Y, Piao X, et al. Traffic data reconstruction via adaptive spatial-temporal correlations[J]. IEEE Transactions on Intelligent Transportation Systems, 2019, 20(4): 1531-1543.

[72]Yu H F, Rao N, Dhillon I S. Temporal regularized matrix factorization for high-dimensional time series prediction[C]. Advances in neural information processing systems, 2016, 847-855.

[73]Takeuchi K, Kashima H, Ueda N. Autoregressive tensor factorization for spatio-temporal predictions[C]. 2017 IEEE International Conference on Data Mining (ICDM). IEEE, 2017, 1105-1110.

Tan H, Feng G, Feng J, et al. A tensor-based method for missing traffic data completion[J]. Transportation Research Part C: Emerging Technologies, 2013, 28(3): 15-27.