DYNAMIC SIMULATION OF A SIMPLE ROUTE GUIDANCE SYSTEM IN THE PRESENCE OF RESPONSIVE SIGNAL POLICIES. IN: BEHAVIORAL AND NETWORK IMPACTS OF DRIVER INFORMATION SYSTEMS

In dynamic, day to day traffic simulation, the issues of route choice is a critical cone. The most common solution is to take a a 'memory based' approach and to assume that drivers 'remember' the costs of travel on a previous day or days. It is then possible for the modeled drivers to prick their routes through the network either by assuming perfect knowledge or by having drivers build up a knowledge base. Conversely, this paper will consider a model of route-guidance system where the drivers routes are chosen based upon information about traffic conditions at that instant but taking no account of the likely evolution of traffic throughout the journey. In a simulation which models a constant level of demand it is likely that the ideal routes through the network will not vary from day to day and therefore it might be expected that a driver routing according to his or her experience of congestion may well have a shorter travel time than a driver who is routing according to an instantaneous estimate of congestion on the current day. This paper investigates this phenomenon and the level of variability of demand which is necessary for a route guidance system based upon instantaneous knowledge to outperform a system where drivers pick their routes wisely based upon knowledge of the network on previous days. This paper also considers the interaction between this route-guidance system and two methods for responsive signal control.