STARIMA for journey time prediction in London.
In: Heydecker, BG, (ed.)
(Proceedings) 5th IMA Conference on Mathematics in Transport.
IMA: Southend on Sea.
Effective prediction of journey time is essential to most transportation management systems and advanced traveller information systems, which allows stakeholders to assess its impact on the road network and enable travellers to effectively reduce the costs of travel time as well. This paper discusses an application of space-time autoregressive integrated moving average (STARIMA) model for journey time prediction. An extended STARIMA model is proposed by considering adjacency relationship of road links as well as the distance of road links when formulating the spatial weight matrix. The extended STARIMA model was tested and a multiple steps prediction is performed using morning peak time dataset in north London. The results show that mean relative error and standard deviation of relative error of extended STARIMA model reduce by 36.36% and 36%, respectively at a prediction horizon of 5 min. The experiment reveals that the extended STARIMA model can obtain the better prediction accuracy than standard STARIMA model in journey time prediction.
|Title:||STARIMA for journey time prediction in London|
|Event:||5th IMA Conference on Mathematics in Transport|
|Dates:||12 April 2010 - 14 April 2010|
|Keywords:||Spatio-temporal analysis, Travel time, Urban congestion|
|UCL classification:||UCL > School of BEAMS
UCL > School of BEAMS > Faculty of Engineering Science
UCL > School of BEAMS > Faculty of the Built Environment
UCL > School of BEAMS > Faculty of the Built Environment > Centre for Advanced Spatial Analysis
Archive Staff Only