setting default latAlignment="left" will give you two queues (left queue for going straight and right queue for turning right). If you want to make it look a bit nicer for your particular scenario you can also use latAlignment="0.5".
Am Mo., 16. Jan. 2023 um 15:19 Uhr schrieb Ni Ying-Chuan < ying-chuan...@ivt.baug.ethz.ch>: > Dear SUMO team, > > > Thanks for the suggestions. > > > I did change the values of *lcSublane* and *lcSpeedGain*. It turned out > that we could indeed prevent the blocking situation by doing so, but the > outcome was that basically all bicycles queue on the right side of the > lane, which is unrealistic. Is it possible to split the flow into two types > (turning and non-turning) and only change the values for those bicycles in > the turning flow? (Namely, applying different junction turning > probabilities to different flows) I found it difficult to achieve when > using jtrrouter. > > > Attached please find the files. > > > The second suggestion you gave is also nice and practical, but it may make > the network more complicated. I will consider that if simply changing the > values fails to achieve what we want. > > > Thank you very much. > > > Best regards, > > > ------------------------------ > *From:* sumo-user <sumo-user-boun...@eclipse.org> on behalf of Jakob > Erdmann <namdre.s...@gmail.com> > *Sent:* Tuesday, 10 January 2023 8:55:55 AM > *To:* Sumo project User discussions > *Subject:* Re: [sumo-user] Turning movement on a sublane > > Hi, it should work by increasing the value of lcSublane and possibly > lowering the value of lcSpeedGain. If not, please attach the scenario files. > Also, you might actually split the single bicycle lane into several narrow > lanes at the junction to define dedicated queues for each turning movement > (see image) > [image: grafik.png] > > > Am Di., 10. Jan. 2023 um 03:03 Uhr schrieb Ni Ying-Chuan < > ying-chuan...@ivt.baug.ethz.ch>: > >> Hi, >> >> >> I am using the sublane model to simulate bicycle traffic on a bike lane >> network with signals. The lane width is 2 m, which is divided by four 0.5 >> m-width sublanes. The *latAlignment* is set to "right". >> >> >> When I included turning movements at intersections, I observed that some >> bicycles would queue on the left side of the lane and could not find the >> proper gap to switch back to the right side and make a right-turning >> movement during green light because of other straight bicycles on the right >> side of the lane. It hence blocked the bicycles behind it a little bit and >> reduced the flow rate. >> >> >> I tried to tune several parameters in the SL2015 model to avoid this. For >> instance, I set *lcTurnAlignmentDistance* to 100, but this would not >> make any difference since the *latAlignment* is on the right already. I >> also tuned *lcSublane*, *lcPushy*, and *lcAssertive* to encourage or >> enable bicycles to stay on the right, but some turning bicycles still go to >> the left when there is space for queueing. Is there any way to ensure >> these turning bicycles always keep themselves on the right side of the lane >> instead of queueing on the left in the sublane model setup? >> >> >> Thank you very much for answering me! >> >> >> _______________________________________________ >> sumo-user mailing list >> sumo-user@eclipse.org >> To unsubscribe from this list, visit >> https://www.eclipse.org/mailman/listinfo/sumo-user >> > _______________________________________________ > sumo-user mailing list > sumo-user@eclipse.org > To unsubscribe from this list, visit > https://www.eclipse.org/mailman/listinfo/sumo-user >
_______________________________________________ sumo-user mailing list sumo-user@eclipse.org To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/sumo-user
