In this study the modelling framework of static traffic assignment is extended to the case where random arc costs are jointly distributed, although implicitly through a local correlation structure at nodes. The resulting network loading problem is then solved through a Monte Carlo simulation a teach node processed in topological order. This way of reproducing the day-by-day randomness of transport supply makes it possible to highlight the role of en-route information to drivers, who are intended to make adaptive choices. The two limit situations where maximum level of information is supplied by a route guidance system and minimum level is self-obtained are considered. For the covering abstract see ITRD E144727. Reprinted with permission of Elsevier
Abstract