Physically Sound Vehicle-Driver Model for Realistic Microscopic Simulation

As far as time is concerned, traffic simulator can be discrete or continuous, depending on whether the state of the system is sampled at discrete or continuous intervals. According to the level of detail requested by the simulation, approaches are in general classified as either macroscopic, mesoscopic or microscopic (Tango, 1997) In microscopic traffic simulators, the interaction between vehicles is considered. The dynamics of the system then becomes an emergent property of the interaction between its components. Such interactions include lane-changing, car-following behaviours and intersection management. Although, in some cases, such simulators produce realistic prediction, they do not allow for an important aspect of the dynamics, that is, the driver-vehicle interaction. Caccibue et al. are investigating driver-vehicle-environment (DVE) interaction models that can predict the human behaviour under different traffic scenarios. According to the authors, such models will prevent erroneous and risky maneuvers, as well as implement means of prompt intervention. The Adaptive Integrated Driver-Vehicle Interface (AIDE) project has been set up to realize the DVE concepts in order to build safer road vehicles. In this paper, a physically sound vehicle-driver model for realistic microscopic simulation, grounded on the DVE paradigm, is introduced. Such a new scheme will enhance accuracy of current traffic models, thus broadening the scope within which traffic simulation can take place.