Journey time forecasting for dynamic route guidance systems in incident conditions

Hounsell, Nick B. and Ishtiaq, Saeed (1997) Journey time forecasting for dynamic route guidance systems in incident conditions. International Journal of Forecasting, 13 (1), 33-42. (doi:10.1016/S0169-2070(96)00698-X).

Abstract

New in-vehicle systems for route guidance require optimum routes in a network to be calculated based on current and forecast journey times. Following a brief review of forecasting methods developed for normal traffic conditions, this article describes a new method for the more difficult but particularly important situation of traffic incidents which occur in variety of forms in urban networks, e.g. an accident, a vehicle breakdown, illegal parking/ stopping and so on. In such conditions journey times may be increased not only on the incident link, but also on the links which are the upstream links of the incident location, this could lead to serious congestion, a rise in energy consumption and environmental nuisance.

The prediction of the effects of traffic incidents is therefore an important issue for better efficiency and for on-line dynamic route guidance (DRG) systems and other traffic control systems. In this study an incident data base was compiled, based on modelling of several incident/ network/ traffic scenarios using a simulation tool. Generalised statistical models were then developed for predicting the spread of congestion effects following an incident and the required travel time modifications on the incident link and on affected links. The aim was to provide a reasonably robust process for on-line applications, to improve on current ad-hoc methods. The main application of the developed models is in incident management for dynamic route guidance systems particularly in low penetration level (i.e. where the proportion of guided drivers is relatively low). New in-vehicle systems for route guidance require optimum routes in a network to be calculated based on current and forecast journey times.

Following a brief review of forecasting methods developed for normal traffic conditions, this article describes a new method for the more difficult but particularly important situation of traffic incidents which occur in variety of forms in urban networks, e.g. an accident, a vehicle breakdown, illegal parking/ stopping and so on. In such conditions journey times may be increased not only on the incident link, but also on the links which are the upstream links of the incident location, this could lead to serious congestion, a rise in energy consumption and environmental nuisance. The prediction of the effects of traffic incidents is therefore an important issue for better efficiency and for on-line dynamic route guidance (DRG) systems and other traffic control systems. In this study an incident data base was compiled, based on modelling of several incident/ network/ traffic scenarios using a simulation tool. Generalised statistical models were then developed for predicting the spread of congestion effects following an incident and the required travel time modifications on the incident link and on affected links.

The aim was to provide a reasonably robust process for on-line applications, to improve on current ad-hoc methods. The main application of the developed models is in incident management for dynamic route guidance systems particularly in low penetration level (i.e. where the proportion of guided drivers is relatively low).

This record has no associated files available for download.

Contributors

Download statistics