Microscopic and mesoscopic traffic models

Besides macroscopic traffic flow models, traffic modelling in freeway systems has also been treated with other general approaches, resulting in microscopic and mesoscopic models. Macroscopic models can surely represent large networks efficiently, since they adopt an aggregate representation of the traffic dynamics, but they generally lack the level of detail needed in modelling the individual drivers’ behaviours and choices. Microscopic models are, instead, conceived to explicitly reproduce the drivers’ responses to traffic patterns, reactions to traffic variations, interactions with other vehicles and route choices, i.e. most of the individual behaviours. Consequently, microscopic models are able to provide a lot of information about the features of traffic flow but they require a high computational effort, especially for large road networks. Mesoscopic models fill the gap between microscopic and macroscopic models, by representing the choices of individual drivers at a probabilistic level, but limiting the level of detail on driving behaviours.