Rolling horizon predictions of bus trajectories

Rolling horizon predictions of bus trajectories

Oshyani, M.F.
Cats, O.

Civil Engineering and Geosciences

Transport & Planning

2014-06-04

Bus travel times are subject to inherent and recurrent uncertainties. A real-time prediction scheme regarding how the transit system evolves will potentially facilitate more adaptive operations as well as more adaptive passengers’ decisions. This scheme should be tractable, sufficiently fast and reliable to be used in real time applications. For this purpose, a heuristic hybrid scheme for departure time estimation is proposed in this study. The predic-tion generated by the proposed hybrid scheme consists of three travel time components: schedule, instantaneous and historical data sources. Genetic algorithm is applied in order to specify the contribution of each data source component to the prediction scheme. The pro-posed scheme was applied for a trunk bus line in Stockholm, Sweden. In addition, the current-ly deployed scheme was replicated in order to compare the performance of both schemes. The results suggest that the proposed scheme reduces the overall mean absolute error by almost 20%. Moreover the proposed scheme provides better predictions except for very long term predictions where both schemes yield the same performance.

prediction
bus departure time
optimization
travel time and genetic algorithm

http://resolver.tudelft.nl/uuid:cb6544e8-02f9-403c-8540-698b7af9a185

National Technical University of Athens (NTUA)

978-960-99994-6-5

OPT-i: Proceedings of the 1st International Conference on Engineering and Applied Sciences Optimization, Kos, Greece, 4-6 June 2014

Institutional Repository

conference paper

(c) 2014 The Author(s)