基于时间序列法的公交车站间行程时间预测模型研究——以苏州1路公交为例
|
摘要
准确预测公交站间行程时间是先进的出行者信息系统(ATIS)的核心,是公交运营和管理的关键问题之一。本文采用改进的时间序列法进行公交车辆站间行程时间预测,利用游程检验法检验时间序列平稳性,并借鉴EMD算法把非平稳时序分解为若干个平稳时序的线形组合,然后对每个平稳时序用指数平滑模型预测。最后,依托苏州1路公交运行数据对该预测模型进行验证,结果表明,该预测模型不仅能够准确的预测公交站间行程时间,还能增强时间序列法对于突发事件的反应能力,大幅提高预测精度。
The effective prediction of bus inter-site travel time is the key information of Advanced Traveler Information System(ATIS) and is crucial to bus operation and management. The paper applies modified time series method to predict bus inter-site travel time. Run Test is used to test the stability of times series and EMD algorithm is applied to decompose the unstable time series into piecewise linear time series. Index smoothing model is then used to predict each time series. The forecasting model is tested based on real-world data collected for the No. 1 bus route in the city of Suzhou. Results indicate that themodel can accurately predict bus inter-site travel time and yield robust prediction under unexpected events.
The effective prediction of bus inter-site travel time is the key information of Advanced Traveler Information System(ATIS) and is crucial to bus operation and management. The paper applies modified time series method to predict bus inter-site travel time. Run Test is used to test the stability of times series and EMD algorithm is applied to decompose the unstable time series into piecewise linear time series. Index smoothing model is then used to predict each time series. The forecasting model is tested based on real-world data collected for the No. 1 bus route in the city of Suzhou. Results indicate that themodel can accurately predict bus inter-site travel time and yield robust prediction under unexpected events.
引文
[1]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.
[2]李天雷.基于GPS数据的公交行程时间计算与预测系统[D].长春:吉林大学,2009.
[3]周文霞,徐建闽,刘正东.基于卡尔曼滤波算法的公交车辆行程时间预测[J].交通标准化,2007(2):174-177.
[4]孙文飞,杨庆芳.基于广义神经网络的公交车行程时间预测方法[A].中国城市规划学会城市交通规划学术委员会、福州市人民政府.公交优先与缓堵对策——中国城市交通规划2012年年会暨第26次学术研讨会论文集[C].中国城市规划学会城市交通规划学术委员会、福州市人民政府,2012:5.
[5]陈旭梅,龚辉波,王景楠.基于SVM和Kalman滤波的BRT行程时间预测模型研究[J].交通运输系统工程与信息,2012,12(4):29-34.
[6]张景.基于定位数据分析的公交时刻表的编制[D].武汉:华中科技大学,2013.
[7]朱丽颖.公交车辆行程时间预测方法研究[D].北京:北京交通大学,2010.
[8]林树宽,杨玫,乔建忠,等.一种非线性非平稳时间序列预测建模方法[J].东北大学学报:自然科学版,2007,28(3):325-328.
[9]刘罗曼.时间序列平稳性检验[J].沈阳师范大学学报:自然科学版,2010,28(3):357-359.
[10]徐大江.预测模型参数的指数平滑估计法及其应用的进一步研究[J].系统工程理论与实践,1999,19(2):26-31,44.
[11]吴小华.Excel在指数平滑法参数优选中的应用[J].安徽工业大学学报:社会科学版,2007,24(1):38-39.
[12]Huang N E,Shen Z,Long S R,et al.The empirical model decomposition and the Hilbert spectrum for non-linear and non-stationary time series analysis[J].Proceedings of the Royal Society of London:Series A,1998(454):903-995.
[2]李天雷.基于GPS数据的公交行程时间计算与预测系统[D].长春:吉林大学,2009.
[3]周文霞,徐建闽,刘正东.基于卡尔曼滤波算法的公交车辆行程时间预测[J].交通标准化,2007(2):174-177.
[4]孙文飞,杨庆芳.基于广义神经网络的公交车行程时间预测方法[A].中国城市规划学会城市交通规划学术委员会、福州市人民政府.公交优先与缓堵对策——中国城市交通规划2012年年会暨第26次学术研讨会论文集[C].中国城市规划学会城市交通规划学术委员会、福州市人民政府,2012:5.
[5]陈旭梅,龚辉波,王景楠.基于SVM和Kalman滤波的BRT行程时间预测模型研究[J].交通运输系统工程与信息,2012,12(4):29-34.
[6]张景.基于定位数据分析的公交时刻表的编制[D].武汉:华中科技大学,2013.
[7]朱丽颖.公交车辆行程时间预测方法研究[D].北京:北京交通大学,2010.
[8]林树宽,杨玫,乔建忠,等.一种非线性非平稳时间序列预测建模方法[J].东北大学学报:自然科学版,2007,28(3):325-328.
[9]刘罗曼.时间序列平稳性检验[J].沈阳师范大学学报:自然科学版,2010,28(3):357-359.
[10]徐大江.预测模型参数的指数平滑估计法及其应用的进一步研究[J].系统工程理论与实践,1999,19(2):26-31,44.
[11]吴小华.Excel在指数平滑法参数优选中的应用[J].安徽工业大学学报:社会科学版,2007,24(1):38-39.
[12]Huang N E,Shen Z,Long S R,et al.The empirical model decomposition and the Hilbert spectrum for non-linear and non-stationary time series analysis[J].Proceedings of the Royal Society of London:Series A,1998(454):903-995.