DRACULA: Dynamic Route Assignment Combining User Learning and Microsimulation

Abstract

Traditionally the analysis of traffic in urban road networks is based on the concept of equilibrium whereby a fixed trip matrix for the design period to be evaluated is assigned to a network where travel times on each link can be defined precisely by a monotonically increasing cost-flow function. Under equilibrium conditions the choice of route or routes is governed by the Wardrop principle that all routes used have equal and minimum travel costs themselves being determined through the above mentioned cost-flow relationships. Thus we are making very precise assumptions not only as to how many drivers travel each O-D pair but also as to which routes they use (although not necessarily the exact proportions) and what the consequent travel times will be. In addition temporal variations within the time period are generally ignored. Clearly this picture represents an over-simplification of real life . The number of vehicle trips per O-D pair varies between days as indeed do the individual drivers who make up those trips. Equally travel conditions vary widely from day to day, partly due to fluctuating demand but also due to factors such as weather conditions vary incidents etc. And finally not all drivers succeed in finding the optimum route, particularly for infrequent drivers. There is clear evidence that the net effect of variability in supply and demand conditions is to significantly increase the mean values of traffic outputs such as traffic time and fuel consumption. For example Mutale (1992) found a 14% increase in travel times over equilibrium due to variability in a north Leeds network. In addition in recent years there has been a massive increase in " real time " advanced technology solutions which both influence and respond to traffic behaviour . At the network wide level these include: