Hi Yu,
you are completely right. Thank you for looking closely into this part of SUMO. The induction loop detector length feature has been retrofitted last year but the speed formula hadn't been updated. This has been corrected and you can either build fresh SUMO binaries from Github source or wait for tomorrow's development binaries to try it out.
Best regards
Mirko
-----Original-Nachricht-----
Betreff: [sumo-user] Another possibly severe problem of induction loops in SUMO
Datum: 2023-07-27T06:12:05+0200
Von: "Ludwig Tang via sumo-user" <sumo-user@xxxxxxxxxxx>
An: "Sumo project User discussions" <sumo-user@xxxxxxxxxxx>
Hi all,
I have one concern about speed data from induction loops. In the default setting, SUMO set the length of induction loops as 0 meter, while in practice the length usually equals around 2 meters. I find that in SUMO induction loop lengths could significantly affect speed measurements. Admittedly, the length may have impacts on observations. I expect the impacts are minor, but it seems not in SUMO. Attached are my experiment results. I considered free-flow traffic on one lane. Clearly, the flow measurements are similar, but the speed data significantly drop with an increase of length of induction loop sensor. I can't understand this phenomenon. Can anyone explain this? I also attached a zip file containing the simulation files and the code.
Best wishes,
Yu