For creating the FD, see https://sumo.dlr.de/docs/Tutorials/FundamentalDiagram.html Regarding your general validation question, you should first figure out what kind of ground truth data you can get. Obtaining the same kind of data from a simulation is usually the easy part.
Am Mo., 8. Juni 2020 um 15:26 Uhr schrieb Tetris < [email protected]>: > Hello, > > I thought it was interesting to compare variations of my simulation > (various > amounts of vehicles) using macroscopic fundamental diagrams. But I am not > sure about the workflow. > > https://sumo.dlr.de/docs/Tutorials/FundamentalDiagram.html > <https://sumo.dlr.de/docs/Tutorials/FundamentalDiagram.html> > > If I want to investigate one hour of a simulation, should the freq of > edgeData output be set to 3600? > Or should freq be set to 60 or 1? And the next step would be to get > "density", "entered" (flow) and "speed" with a script from the xml to excel > and then into matlab to plot the fundamental diagram? Every edge of the > network should have a density, entered and speed column itself? Or should I > take every value of the interval and calculate the average for every > interval (eg one second or one minute of the simulation)? > > Is there an easier way to get the fundamental diagram? > > Is it also meaningful to get a fundamental diagram of just one edge? Or the > average speed of all vehicles in the simulation? What I originally wanted > to > achieve is evidence of good applicability to the real world and find out > which simulation is the best. I already compared the travel time and speed > for one vehicle. Is it adequate to take a look at just one vehicle (which > is > of course influenced by all the other vehicles) to validate the simulation > or should the values of all the other vehicles be taken into account too? > > > > -- > Sent from: http://sumo-user-mailing-list.90755.n8.nabble.com/ > _______________________________________________ > sumo-user mailing list > [email protected] > To unsubscribe from this list, visit > https://www.eclipse.org/mailman/listinfo/sumo-user >
_______________________________________________ sumo-user mailing list [email protected] To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user
