Int. J. of Computational Intelligence Studies   »   2017 Vol.6, No.4

 

 

Title: Portfolio theory application to prediction correction of train arrival times

 

Authors: Takaaki Yamada; Tatsuhiro Sato

 

Addresses:
Center for Technology Innovation – Systems Engineering, Hitachi, Ltd., Yokohama, Japan
Center for Technology Innovation – Systems Engineering, Hitachi, Ltd., Yokohama, Japan

 

Abstract: The application of portfolio theory to the prediction of train arrival times is shown to improve prediction accuracy. The 'portfolio' comprises two correction methods based on a Wiener process: one uses history data for the current day and the other uses history data for previous days. These data represent several properties such as station and are used after being statistically cleaned. The error between the time predicted using a local model and the actual time is assumed to have a normal distribution. Portfolio theory is used to determine the optimal application of the two methods to the correction process. Simulation using actual dense train schedules showed that the average error in the predicted arrival time was reduced from 12 s to 3 s. This error reduction will, for example, improve the efficiency of regenerative braking systems, in which the kinetic energy of an arriving train is electrically transmitted to a departing train.

 

Keywords: railway; train arrival time; prediction correction; prediction accuracy; portfolio theory; portfolio optimisation.

 

DOI: 10.1504/IJCISTUDIES.2017.10010615

 

Int. J. of Computational Intelligence Studies, 2017 Vol.6, No.4, pp.349 - 365

 

Submission date: 16 Dec 2016
Date of acceptance: 10 Aug 2017
Available online: 23 Jan 2018

 

 

Editors Full text accessAccess for SubscribersPurchase this articleComment on this article