Genaro Peque, Jr | Toshihiko Miyagi | Fumitaka Kurauchi
OPEN ACCESS
A multi-agent route choice learning model for the microscopic simulation-based dynamic traffic assignment (DTA) is used to investigate the effects of traffic information accuracy on drivers' day-to-day route choice decisions. Using the total relative gap convergence metric to quantify the convergence speed for some chosen update cycle length intervals, the results show that a slight decrease in accuracy has a negative effect on the rate of convergence. From a learning perspective, shorter information up-date cycles from an advanced traveller information system induce faster convergence when compared to longer information update cycles. This implies that drivers learn faster, given the additional computational and storage costs of travel information that the system is willing to invest in. Moreover, when the update cycle length is very long, it produces a worse result compared to a scenario where drivers rely only on their own travel experiences based on the routes they have chosen.
ATIS, day-to-day route choice, simulation-based DTA
[1] Avineri, E. & Prashker, J.N., The impact of travel time information on travelers' learning under uncertainty. Transportation, 33, pp. 393–408, 2006. DOI: 10.1016/0191-2615(92)90017-Q.
[2] Ben-Elia, E., Erev, I. & Shiftan, Y., The combined effect of information and experience on drivers route-choice behavior. Transportation, 35, pp. 165–277, 2008. DOI: 10.1037/h0092987.
[3] Bogers, E.A.I. & van Zuylen, H.J., The influence of reliability on route preferences in freight transport. Paper Presented to the 84th Annual Meeting of the Transportation Research Board, Washington and published in Transportation Research Record, 2005.
[4] Ben-Elia, E., Di Pace, R., Bifulco, G.N. & Shiftan, Y., The impact of travel information's accuracy on route-choice. Transportation Research Part C, 26, pp. 146–159, 2013. DOI: 10.1016/j.trc.2012.07.001.
[5] Cascetta, E., Transportation Systems Engineering: Theory and, Methods, Kluwer Academic Publishers: Dordrecht, 2001.
[6] Prashker, J. & Bekhor, S., Route-choice models used in the stochastic user equilibrium problem: a review. Transport Reviews, 24, 437–463, 2004. DOI: 10.1016/S0191-2615(00)00045-X.
[7] Miyagi, T., Multiagent learning models for route choices in transportation networks: an integrated approach of regret-based strategy and reinforcement learning. Proceedings of the 11th International Conference on Travel Behavior Research, Kyoto, 2006.
[8] Miyagi, T. & Peque, G., Informed user algorithm that converge to a pure Nash equilibrium in traffic games. Procedia – Social and Behavioral Sciences, 54, pp. 438–449, 2012. DOI: 10.1016/j.sbspro.2012.09.762.
[9] Peque, G., Miyagi, T. & Kurauchi, F., Adaptive learning algorithms for simulationbased dynamic traffic user equilibrium. International Journal of Intelligent Transportation Systems Research, 16(3), pp. 215–226, 2018. DOI: 10.1023/A:1007678930559.
[10] Miyagi, T., Peque, G. & Fukumoto, J., Adaptive learning algorithms for traffic games with naive users. Procedia – Social and Behavioral Sciences, 80, pp. 806–817, 2013. DOI: 10.1016/j.sbspro.2013.05.043.
[11] Selten, R., Schreckenberg, M., Chmura, T., Pitz, T., Kube, S., Hafstein, S., Chrobok, R., Pottmeier, A. & Wahle, J., Experimental investigation of day-to-day route-choice behaviour and network simulations of autobahn traffic in North Rhine-Westphalia. Human Behaviour and Traffic Networks. eds. A. Schreckenberg & R. Selten, Springer: Berlin Heidelberg, pp. 1–21, 2004.
[12] Horowitz, J.L., The stability of stochastic equilibrium in a two-link transportation network. Transportation Research Part B: Methodological, 18, pp. 13–28, 1984. DOI: 10.1016/0191-2615(84)90003-1.
[13] Daganzo, C. & Sheffi Y., On stochastic models of traffic assignment. Transportation Science, 11, 253–274, 1977. DOI: 10.1287/trsc.11.3.253.
[14] Sheffi, Y. & Powell, W., An algorithm for the equilibrium assignment problem with random link times. Networks, 12, 191–207, 1982. DOI: 10.1016/0191-2615(81)90046-1.
[15] Dynamic Traffic Assignment: A Primer, Transportation Network Modeling Committee, pp. 11, 2011.
[16] Singh, S., Jaakkola, T., Szepesvari, C. & Littman, M., Convergence results for singlestep on-policy reinforcement-learning algorithms. Machine Learning 38(3), pp. 287– 308, 2000. DOI: 10.1023/A:1007678930559.
[17] Leslie, D. & Collins, E., Generalised weakened fictitious play. Games and Economic Behaviour, 56, pp. 285–298, 2006. DOI: 10.1016/j.geb.2005.08.005.
[18] Marden, J., Young, P., Arslan, G. & Shamma, J., Payoff-based dynamics for multiplayer weakly acyclic games. SIAM Journal on Control and Optimization, 48(1), 2009. DOI: 10.1137/070680199.
[19] Chapman, A., Leslie, D., Rogers, A. & Jennings, N., Convergent learning algorithms for unknown reward games. SIAM Journal on Control and Optimization 51(4), pp. 3154–3180, 2013. DOI: 10.1137/120893501.
[20] Krauß, S., Towards a unified view of microscopic traffic flow theories. Proceedings of the 8th IFAC/IFIP/IFORS Symposium, Chania, Greece, 16–18 June, eds. M. Papageorgiou & A. Pouliezos, Vol. 2, Elsevier Science, 1997.
[21] Nagel, K. & Schreckenberg, M., A cellular automaton model for traffic flow. Journal de Physique I, 2, p. 2221, 1992. DOI: 10.1051/jp1:1992277.