Mesoscopic simulation uses a simplified car following model based on the work of M. Mahut (References). The main difference to car following models used in microscopic simulation is that braking and accelerating are not included. A vehicle changes its speed immediately and has only two statuses:
The precise formulation of a car following model depends on the meso speed model selected. You may select a meso speed model for each link (Attributes of links):
The car following model uses fixed link speeds:
These include:
xFollower: Front end position of trailing vehicle on current link
xLeader: Front end position of preceding vehicle on current link
VLink: Meso link speed
τFollower: Response time of trailing vehicle
λLeader: Effective vehicle length of preceding vehicle, calculated from the individual vehicle length and the meso standstill distance.
The car following model uses individual vehicle speeds:
In this formulation, VLink is replaced by VFollower: Each vehicle tries to drive at its individual desired speed VFollower.
Both car following models use individual vehicle lengths and response times that are based on driving behavior parameters.
By default, the meso speed model is set to Vehicle-based. Vehicles therefore drive at their individual desired speed in free flowing traffic. Desired speed decisions and reduced speed areas result in changes to the vehicle’s desired speed. When used the following limitations apply:
For the Link-based meso speed model the following applies:
Desired speed decisions and reduced speed areas are ignored. Differences in the desired speed of vehicle classes cannot be modeled. This means that different speed limits on different lanes can only be modeled with the vehicle-based model. However, in a city center network, the Link-based meso speed model might be sufficient. Use of the Link-based meso speed model in most parts of the network can lead to enhanced simulation speed.
