faq 170688515

Is 'maximumVelocity' ignored in 'equil-mixedTraffic' example?

by Wolfgang Schieder on 2018-02-07 10:50:11

---

Hello together,

I ran the 'equil-mixedTraffic' with MATSim 0.9.0 (for details see attached ZIP-File containing the output directory) and try to understand the 'output_plans.xml' regarding the velocity of 'car' and 'bicycle' mode:

The 'freespeed' in the 'network.xml' is 27.78 m/s.

The 'maximumVelocity' in the 'vehicles.xml' is 16.67 m/s for 'car' mode.

The 'maximumVelocity' in the 'vehicles.xml' is 4.17 m/s for 'bicycle' mode.

Travel times of persons with ID=1 and ID=10 (first plans) seem to be the same for 'car' and 'bicycle' mode on the identic route (00:08:59 and 00:32:59 travelling distances of 25 000 m and 65 000 m). I calculated the speed out of this given information approximately at 46 m/s and 33 m/s for both, 'car' and 'bicycle' mode.

This exceeds 'maximumVelocity' in 'vehicles.xml' and 'freespeed' in 'network.xml' and is not plausible, as 'car' and 'bicycle' seem to have the same speed.

Any suggestions, how to interpret those results?

Could I have the units get mixed up, however, 'bicycle' mode should be slower than 'car' mode anyway?

Is 'maximumVelocity' ignored in 'equil-mixedTraffic' example?

Thanks in advance,

Wolfgang

---

Comments: 2

---

Re: Is 'maximumVelocity' ignored in 'equil-mixedTraffic' example?

by Kai Nagel on 2018-02-27 20:41:25

Actually, looking at output_plans, or <iteration>.plans, is rather useless for the question here, since most of the times in the plans file are never used when the modes are executed on the network. The first activity ends at its endTime from the plans file, the trip is executed on the network by the mobsim where thus the travel time is what it takes, the start of the next activity is thus also just given by the arrival time, and then it takes either the duration or the endTime of the activity. In these situations, travelTime and activityStartTime are informational only, and may be wrong. (The travelTime comes, as an estimate, from the router, and it is indeed a bit of a question why they are so far of. But it does not affect the validity of the simulation.)

Also see https://matsim.atlassian.net/wiki/display/CQ/customcontent/189235210 .

Quite in general, analysis should be based on output_events . Everything else is for information only.

---

Re: Is 'maximumVelocity' ignored in 'equil-mixedTraffic' example?

by Amit Agarwal on 2018-02-27 21:54:00

Hello Wolfgang,

Thanks for the pointing the problem with enough details.

1. First of all, as Kai already said, output_plans are not useful to analyse the simulation results. You should rather use output_events which represents the simulation outcome.

2. In the output events, the travel time on link 21 for car (agent 1) and bicycle (agent 10) are 600 sec and 2399 sec respectively. Length of the link 21 is 10km. Thus, the speeds of car and bicycle modes are 16.67 and 4.17 m/s. In short, the queue model considers link free speed as well as the maximum speed of the vehicle.

3. Your observation in the plans file is correct. In absence of the mode-specific network router,

   for all modes, car network router is used. It is possible to have mode-specific network router in which one can incorporate the maximum speed of the vehicle, however, for simplicity, the mode-specific network router is not used in this example. Therefore, the travel times in the plans are same.

   0.9.0 already have the mode-specific routers. As Kai says in comment below, same travel times are due to non-consideration of maximum vehicle speed in the router.

4. Lastly, in general, the link travel times are estimated from link length and link speed. So, I have to check, why link speeds are not honoured in this case. Most likely, the reported distances or the travel times in the plans file are wrong.

I hope, these will help you understand the scenario. Let us know if something is still unclear.