Modelling and Simulation of Mixed Traffic

Modelling and Simulation of Mixed Traffic


Annique Lenorzer, TSS - Transport Simulation Systems S.L., Jordi Casas, TSS - Transport Simulation Systems S.L., Mark Brackstone, TSS - Transport Simulation Systems Ltd.


Modelling of mixed flow traffic and test through a case study.


Traffic simulation is one of the most effective tools for the assessment of road safety, however the simulation of the potentially most conflictive situation like heterogeneous traffic involving vulnerable road users such as bicycles and pedestrians with passenger cars is still in at an early stage.
Recent developments in the modelling of two-wheeled vehicles include the adaptation of a number of existing behavioural models such as Car-following and Lane-changing models, Cellular Automata models (Nagel and Schreckenberg, 1992) and Social Force models (Helbing and Molnar, 1995): The Car-following model can account for each pair of vehicle types present in the heterogeneous flow (Ravishankar and Mathew 2011); or includes Lateral discomfort caused by lateral friction between vehicles (Gunay, 2007). Whereas the problem of dealing with non lane-based lateral movements can be treated either in a discrete form splitting the lanes into narrow strips where vehicles can occupy several strips at a time and applying some sort of lane changing from strip to strip (Mathew et al. 2013).; or in a continuous way by introducing a veering angle, and a path selection process to update the lateral position and lateral speed (Falkenberg et al. 2003; Lee 2008). Similarly to the strip approach for lane-based model, an extension of the cellular automata model with smaller cells and where vehicles can occupy several cells at a time has been proposed by different authors (e.g. Yao et al. 2009, Vasic and Ruskin 2011). Social Force models including different force for each vehicle type pairs (Li et al. 2011), additional contact and friction forces (Liang et al. 2012) have also been proposed.
Although a number of specific models have been proposed to simulate the behaviour of two-wheeled vehicles, the vast majority of them have not been properly validated due to the lack of relevant empirical data.
This paper analyses a modification of behavioural models:
• Car-Following model has been extended to cases where the leader is not aligned with the follower but where a lateral shift is present; as well as cases where a follower has multiple leaders.
• Lane selection model has been extended for considering two dimensions of the space
• Gap-acceptance model has been extended to allow for the inclusion of non-lane based criteria
The analysis of these behavioural models has been complemented with a case study where the behaviour of the mixed traffic is the major component for evaluating the effectiveness of different management strategies.


Association for European Transport