Dynamic and stochastic routing for multimodal transportation systems

Sofie Demeyer, Mario Pickavet, Piet Demeester, Pieter Audenaert
2014 IET Intelligent Transport Systems  
We present a case study of a multimodal routing system that takes into account both dynamic and stochastic travel time information. A multimodal network model is presented that makes it possible to model the travel time information of each transportation mode differently. This travel time information can either be static or dynamic, or either deterministic or stochastic. Next to this, a Dijkstra-based routing algorithm is presented that deals with this variety of travel time information in a
more » ... form way. This research focuses on a practical implementation of the system, which means that a number of assumptions were made, like, for example, the modeling of the stochastic distributions, comparing these distributions, etc. A tradeoff had to be made between the performance of the system and the accuracy of the results. Experiments have shown that our system produces realistic routes in a short amount of time. It is demonstrated that routing dynamically indeed results in a travel time gain in comparison to routing statically. By making use of the additional stochastic travel time information even better (i.e. faster) and more reliable routes can be calculated. Moreover, it is shown that routing in the RELATED WORK The most common (early) routing systems presume the travel time data to be static, i.e. independent of the time of the day. In this way, the well-known algorithm of Dijkstra [1] can be applied in order to find the shortest (i.e., fastest) route between two locations. Since traffic is not a static matter, more accurate travel times can be obtained by taking into account its dynamic character, i.e., the (travel time) cost of a link is dependent on the time of the day this link is
doi:10.1049/iet-its.2012.0065 fatcat:fiwksv3cbzde3ft6mb6gx7rtwq