Hi,
I am working on creating a random road traffic generator, so far i've seen so many random traffic generators either on commercial software or open source software, but i can't seem to figure out how random is it? is there a control on how the arrival generation is done in order to show some phenomenon, like periodicity.
In my case I am working on generating traffic coming from a signalized intersection. while using a distribution with parameters fitted using real data, by generating random Time Headway between vehicles i can't see the periodicity, however when i add a control on how much vehicles should be generated in each period (let's call it the phase in the preceding intersection) using a poisson distribution and after that making sure to not generated Time Headways that would not be more than then the period duration and also not respecting the amount of vehicles generated in a period for computation efficiency (in order to not have the computer do too much loops and take more time in the traffic generation) I can clearly see the periodicity phenomenon. But scientifically speaking is this not interfering with the random generation and adding a bias to my generated data ?
The vehicles leave the signal junction in a platoon with a headway of approximately 2s, the headway varies mostly depending on the environment e.g. nearside or offside lane, lane width curvature gradient wet or dry, there are simple calculations that are generally considered reliable for this (Webster and Cobbe 1966, RR67 TRL).
When the platoon accelerates towards cruising speed it starts to disperse because each driver/vehicle has a different desired cruising speed (normal distribution for similar category of vehicle) and the dispursion increases the further the vehicles travel. There are some more complicated factors though like number of lanes/lane changes, vehicle category proportions etc.
Searching for platoon dispersion modelling yields lots of results :)
Assume Poisson's arrivals at intersections and incorporate a suitable headway distribution. The distributions can be deterministic as discussed in Dr. Webster's comments above.
Thanks,
Dr. Bazlamit
Hello, I need a traffic generator for my research. Can you help me?
I need to generate data using a traffic generator to simulate my work, but I do not know any software. Can you help me solve this problem?
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India