A doubly dynamic models system aiming at simulating the impacts of Drivers Information Systems (DIS) on pre-trip users' travel choices such as route and departure time and on network performances is presented. The overall framework consists of three main sub-models: the Travel Behaviour Simulator, the Traffic Simulator and the DIS Simulator. The Travel Behaviour Simulator models the day-to-day cost and choice updating process for different user classes. The Traffic Simulator simulates the propagation of flows on the network and estimates within-day dynamic network performances allowing for non-stationary conditions within the simulation period and random fluctuation of supply characteristics. Finally, a module generating predictive traffic information simulates the presence of a Drivers Information System (DIS) in the transport system. Supply/Demand interactions are simulated within a day-to-day dynamic process framework. An analysis of expected path travel times for four classes of drivers (given by the cross combination of Informed and Not-Informed, Habitual and Not-Habitual drivers) has been carried on by means of a simple test network, under different condition of congestion (i.e. recurrent and non-recurrent). The models system allows analysing the (individual and global) benefits due to the DIS in the case of predictive not-consistent information. It was observed that the reduction of path travel times is strictly related to the level of market penetration. Since not-consistent information by definition does not take into account the behavioural response of drivers to the information provided, in case of highly congested road networks, the well-known phenomenon of overreaction does occur. The latter can dramatically reduce the benefits of information provision, causing, for certain ranges of market penetration, a worsening of network performances.
Samenvatting