An algorithm for city transport arrival time estimation using adaptive elementary predictions composition
A.A. Agafonov, V.V. Myasnikov

Full text of article: Russian language.

Abstract:
The problem of precise arrival time of public transport is considered in this paper. There is proposed a new prediction algorithm based on adaptive composition model using elementary prediction. A small number of adaptive parameters characterizes each elementary prediction algorithm. Adaptability means that parameters of the constructed compositions depend on a number of control parameters of the model, which includes the following factors: weather conditions, traffic density, driving dynamics, prediction horizon, etc. Adaptability is achieved by introducing a hierarchical decomposition range of control parameters used in regression tree. We made experimental investigations on real routes of city public transport in Samara to evaluate the prediction accuracy of the proposed algorithm. We also explain the advantages of the proposed solution in comparison with existing ones.

Key words:
city public transport, arrival time prediction, arrival time estimation, algorithms composition, hierarchical decomposition, regression tree.

References:

  1. Hall, R. Handbook of transportation science / Randolph W. Hall. – Dordrecht: Kluwer Academic Publishers, 2003. – 737 p.
  2. Altinkaya, M. Urban Bus Arrival Time Prediction: A Review of Computational Models / M. Altinkaya, M. Zontul // International Journal of Recent Technology and Engineering (IJRTE). – 2013. – V. 2, Issue 4. – P. 164-169.
  3. Hoogendoorn, S.P. State-of-the-art of vehicular traffic flow modeling / S.P. Hoogendoorn, P.H.L. Bovy // Proceedings of the Institution of Mechanical Engineers. Part I: Journal of Systems and Control Engineering. – 2001. – V. 215(4). – P. 283-303.
  4. Padmanaban, P. Estimation of Bus Travel Time Incorporating Dwell Time for APTS Applications / R.P.S. Padmanaban, L. Vanajakshi, S.C. Subramanian // IEEE Intelligent Vehicles Symposium. – 2009. – V. 2. - P. 955-959.
  5. Агафонов, А.А. Прогнозирование параметров движения городского пассажирского транспорта по данным спутникового мониторинга / А.А. Агафонов, А.В. Сергеев, А.В. Чернов // Компьютерная оптика. – 2012. – Т. 36, № 3. – С. 453-489.
  6. (Agafonov, A.A. Forecasting of the motion parameters of city transport by satellite monitoring data / A.A. Agafonov, A.V. Sergeyev, A.V. Chernov // Computer Optics. – 2012. – V. 36 (3). – P. 453-458.)
  7. Agafonov, A. City transport motion parameters forecasting by satellite monitoring data and statistics / A. Agafonov, A. Chernov, A. Sergeyev // PRIA-2013. - 2013. - V. 2. – P. 489-491.
  8. Sun, H. Use of Local Linear Regression Model for Short-term Traffic Forecasting / H. Sun, H.X. Liu, H. Xiao, R.R. He, B. Ran // Transportation Research Record. – 2003. – Issue 1836. – P. 143–150.
  9. Vanajakshi, L. Travel time prediction under heterogeneous traffic conditions using global positioning system data from buses / L. Vanajakshi, S.C. Subramanian, R. Sivanandan // IET Intelligent Transport Systems. – 2009. – V. 3. – P. 1-9.
  10. Shalaby, A. Prediction Model of Bus Arrival and Departure Times Using AVL and APC Data / A. Shalaby, A. Farhan // Journal of Public Transportation. – 2004. - V. 7(1). – P. 41-63.
  11. Chen, M. A dynamic bus-arrival time prediction model based on APC data / M. Chen, X. Liu, J. Xia, S.I. Chien // Computer-Aided Civil and Infrastructure Engineering. – 2004. – V. 19(5). – P. 364-376.
  12. Chang, G.-L. Predicting intersection queue with neural network models / G.-L. Chang, C.-C. Su // Transportation Research Part C. – 1995. - V. 3(3). – P. 175-191.
  13. Jeong, R. Bus arrival time prediction using artificial neural network model / R. Jeong, L.R. Rilett // IEEE Conference on Intelligent Transportation Systems, Proceedings, ITSC. – 2004. – P. 988-983.
  14. Bin, Y. Bus arrival time prediction using support vector machines / Y. Bin, Y. Zhongzhen, Y. Baozhen // Journal of Intelligent Transportation Systems: Technology, Planning, and Operations. – 2007. - V. 10, Issue 4. - P. 151-158.
  15. Wu, C.-H. Travel-time prediction with support vector regression / C.-H. Wu, J.-M. Ho, D.T. Lee // IEEE Transactions on Intelligent Transportation Systems. – 2004. – V. 5(4). – P. 276-281.
  16. van Lint, J.W.C. Accurate freeway travel time prediction with state-space neural networks under missing data / J.W.C. van Lint, S.P. Hoogendoorn, H.J. van Zuylen // Transportation Research Part C: Emerging Technologies. – 2005. – V. 13(5-6). – P. 347-369.
  17. Park, T. A bayesian approach for estimating link travel time on urban arterial road network / T. Park, S. Lee // Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). – 2004. – V. 3043. – P. 1017-1025.
  18. Zheng, W. Short-term freeway traffic flow prediction: Bayesian combined neural network approach / W. Zheng, D.-H. Lee, Q. Shi // Journal of Transportation Engineering. – 2006. –V. 132(2). – P. 114-121.
  19. Yang, J.-S. Travel time prediction using the GPS test vehicle and Kalman filtering techniques / J.-S. Yang // Proceedings of the American Control Conference. – 2005. - V. 3. - P. 2128-2133.
  20. Wall, Z. An Algorithm for Predicting the Arrival Time of Mass Transit Vehicles Using Automatic Vehicle Location Data / Z. Wall, D. J. Dailey // 78th Annual Meeting of the Transportation Research Board, Washington D.C., 1999.
  21.  Zaki, M. Online Bus Arrival Time Prediction Using Hybrid Neural Network and Kalman filter Techniques / M. Zaki, I. Ashour, M. Zorkany, B. Hesham // International Journal of Modern Engineering Research. – 2013. – V. 3, Issue 4. – P. 2035-2041.

© 2009, IPSI RAS
Institution of Russian Academy of Sciences, Image Processing Systems Institute of RAS, Russia, 443001, Samara, Molodogvardeyskaya Street 151; e-mail: ko@smr.ru; Phones: +7 (846 2) 332-56-22, Fax: +7 (846 2) 332-56-20