You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
For my case study, I want to run MatSim by using freight contrib alongside with the synthetic population.
My end goal is to calculate the link volume in the study area.
I noticed that we have the vehicles here to generate fleets for passenger cars, it is quite tempting to me to add some stages
to generate the synthetic freight demands, then write out the CarriersPlan.xml and VehiclesTypes.xml via extending matsim.writers.py.
However, I felt a bit short about what public dataset in France I could use and how to generate a realistic one.
Found freightDemandGeneration in matsim, but I found if the demand data is not there, the output will be completely randomized, e.g. the location of depots
Any advices on this? and I guess the eqasim-java is extendable to add addtional contribs as well, right?
The text was updated successfully, but these errors were encountered:
Hi, there are several people working towards this, but it is not really an obvious task. The main source of information is SIRENE and there are surveys from LET that describe for instance generation and reception rates of companies from specific sectors. We have a draft of a methodology that is based on the FRETURB approach, but it is not finished yet. We may look into this in the next months though, and it also depends strongly on the data quality (especially SIRENE) which seems to have deteriorated in recent years, especially during Covid.
If this is interesting to you, we could have a more detailed discussion, and we'd be happy to share what we have and maybe also guide the further development in a collaboration if you would want to invest time on it. But be aware that this is probably a multi-month project in itself, so it depends in which context you are working on that :)
As for the freight contrib in MATSim, I'm not sure what it does and how reliable it is. I only know that it uses JSprit and we have made the experience that it is less performant (runtime and solution quality) than other VRP solvers. But solution quality may not be necessary to get an average idea of the traffic. IMO it would be sufficient to add agents for the freight movements which simply perform rerouting over the iterations and externalize all the VRP-related stuff from the simulation.
For my case study, I want to run MatSim by using freight contrib alongside with the synthetic population.
My end goal is to calculate the link volume in the study area.
I noticed that we have the vehicles here to generate fleets for passenger cars, it is quite tempting to me to add some stages
to generate the synthetic freight demands, then write out the
CarriersPlan.xml
andVehiclesTypes.xml
via extendingmatsim.writers.py
.However, I felt a bit short about what public dataset in France I could use and how to generate a realistic one.
Found freightDemandGeneration in matsim, but I found if the demand data is not there, the output will be completely randomized, e.g. the location of depots
Any advices on this? and I guess the eqasim-java is extendable to add addtional contribs as well, right?
The text was updated successfully, but these errors were encountered: