Motorway Traffic

The assessment of traffic quality on motorways is typically based on macroscopic variables such as travel times and queue lengths. PTV Vissim achieves realistic results at this aggregated level by means of detailed geometry and microscopic behaviour models.

Simulating freeway networks with PTV Vissim: Visualization of link segment speed data.

Operational level: Driving behaviour

Modelling individual driving behaviour forms the core of simulation. With an unlimited number of vehicle models, types, and classes, users can put together any fleet of vehicles required and adjust the driver and vehicle properties in detail. In addition to basic attributes such as the desired speed and acceleration and deceleration behaviour, at the operational level, traffic flow is determined by car following behaviour and lane changing behaviour.

Car following behaviour is simulated using Professor R. Wiedemann's psycho-physical car following model. In this model, the driver responds according to the distance and the difference in speed to the vehicle ahead by accelerating or decelerating. In this way, he adjusts either his desired speed or his safety distance to the vehicle ahead. In PTV Vissim, it is possible to define individual driver characteristics such as the safety distance to the vehicle ahead for different vehicle classes and types.

A rule-based model is used for lane changing, which can also be configured. One example is aggressiveness when changing lanes: what gap size is required in the adjacent lane? And how much deceleration can be accepted for the trailing vehicle on the target lane?

Tactical level: Lane selection and cooperative behaviour

Lane selection is vitally important for the traffic flow: unlike changing lanes, this is not the operational procedure of actually moving over, but rather looking ahead and selecting the desired lane. This primarily depends on the route of the vehicle and the look ahead distance. From this distance before a decision point, a vehicle can try to move to a specific lane. The look ahead can extend over several intersections.

However, lane selection does not just depend on static network characteristics, but also on the surrounding vehicles: Lane selection, acceleration and deceleration allow vehicles to cooperate with other vehicles, for example to facilitate merging at a weaving at a ramp. This in turn has an influence on the capacity of the merging area and is strongly dependent on the situation. Therefore, properties such as the willingness to cooperate can be adjusted in PTV Vissim so that the behaviour in the region being studied can be mapped realistically. Finally, this flexibility also allows the evaluation of traffic management measures and Car2X systems that have an impact on individual behaviour and whose effects on the entire traffic flow should be evaluated.