| Re: [sumo-user] one simple question |
If the routing is enabled for any vehicles, the average travel times on the net are collected for all edges. If a vehicle needs to be routed (either because it gets inserted or because a repeated route choice was enabled via the ".period" option) it chooses the fastest route to its destination edge (or district) according to the present edge weights (travel speeds and hence travel times). The update of the edge weights does not simply overwrite the old value, but is computed as a weighted average as described below. Since updating the weights of all edges in each simulation step means a major slowdown for the simulation this interval may be altered using the ".adaptation-interval" option.
***************************************************************************************************************************************************************************************************************************************************************************************************
All the commands below are mentioned above determine the travel speeds which give to SUMO to take as as reference to simulate the travel time . The travel time will then be got by summing every edge length divided by the travel speed for each edge mentioned above.
IS THAT RIGHT?
Best!
Jane
Jakob, thanks so much!!After reading the links you attached, I still would like to ask a few quesitons and Thanks in advance:Q2, Q4: no. Sumo will load vehicles that depart within the next 300 seconds. This will probably not be noticable in execution time unless you have very many vehicles. Then after having simulation ~300 seconds it will load vehicles for the next 300 seconds.This is faster than loading new vehicles in every simulation step and conserves memory compared to loading all at once.I got it!!The route computation only depends on edge length and edge travel time. There multiple algorithsm/options for estimating this travel time: http://sumo.dlr.de/wiki/Simulation/Routing#Travel-time_values_for_routingQuestion:Does it mean that traffic lights are also an indirect factor that affects the rerouting route choice, in other word, the traffic lights control model are considered in the rerouting algorithm by SUMO. Because the traffic light control model affects the edge travel time?if the command "-device.rerouting.probability=xx" was NOT defined in xx.sumocfgWill SUMO defaults that "--device.rerouting.pre-period =60" even though we did not write this command in the xx.Yes. The default is always active. There are other ways of giving vehicles the device (via other options, xml parametesr or traci calls)I got it ! Thanks!-device.rerouting.adaptation-stepsThis option actives an alternative algorithm for averaging travel times: http://sumo.dlr.de/wiki/Demand/Automatic_Routing#Adapting_by_moving_averageQuestion:
--device.rerouting.adaptation-interval <INT> 1 The interval for updating the edge weights. --device.rerouting.adaptation-weight <FLOAT> 0.5 The weight of prior edge weights for exponential averaging from [0, 1]. --device.rerouting.adaptation-steps <INT> 0 The number of adaptation steps for averaging (enable for values > 0). in my understanding, the weight is associated the edge speed , what's more, the weight or the edge speed is affected by the steps (which affect the previous average speed value) and the interval(affect the current speed)for example, one simulation will last about 900s, if we defined these commands,--device.rerouting.adaptation-interval = 1(means1 second )--device.rerouting.adaptation-weight =0--device.rerouting.adaptation-steps=3When the simulation is running at the time of 600s,For the formulation1) Float=0 (-device. rerouting. adaptation-weight =0) which means DO NOT take the previous speed information into the consideration2) Prior value=the last three (--device.rerouting.adaptation-steps=((meanspeed in 599s-600s) +(,meanspeed in 598s-599s)+(meanspeed in 597s-598s))/3For 599s-600s, 598s-599s, 597s-598s is determined by the --device.rerouting.adaptation-interval = 1If we set --device. rerouting. adaptation-interval = 2 we need the mean speed of 598s-600s 596s-598s 594s-596sOn Thu, Sep 13, 2018 at 1:23 AM Jakob Erdmann <namdre.sumo@xxxxxxxxx> wrote:Q2, Q4: no. Sumo will load vehicles that depart within the next 300 seconds. This will probably not be noticable in execution time unless you have very many vehicles. Then after having simulation ~300 seconds it will load vehicles for the next 300 seconds.This is faster than loading new vehicles in every simulation step and conserves memory compared to loading all at once.The route computation only depends on edge length and edge travel time. There multiple algorithsm/options for estimating this travel time: http://sumo.dlr.de/wiki/Simulation/Routing#Travel-time_values_for_routingif the command "-device.rerouting.probability=xx" was NOT defined in xx.sumocfgWill SUMO defaults that "--device.rerouting.pre-period =60" even though we did not write this command in the xx.Yes. The default is always active. There are other ways of giving vehicles the device (via other options, xml parametesr or traci calls)-device.rerouting.adaptation-stepsThis option actives an alternative algorithm for averaging travel times: http://sumo.dlr.de/wiki/Demand/Automatic_Routing#Adapting_by_moving_average2018-09-13 1:31 GMT+02:00 Jane Cheung <cheung180515@xxxxxxxxx>:
--device.rerouting.adaptation-steps <INT> 0 The number of adaptation steps for averaging (enable for values > 0). for example
--device.rerouting.adaptation-steps =3 Does this command is used to limit the frequency (or number of times) that one vehicle could choose to reroute the route for itself, even though (--device. rerouting. Period /probability/deterministic/pre-period) have been defined?In example, it means one or all vehicles could only choose to reroute 3 times at most?Jakob, Thanks a lot!Looking forward for your reply!Best!JaneOn Wed, Sep 12, 2018 at 3:58 PM Jakob Erdmann <namdre.sumo@xxxxxxxxx> wrote:Q1: they are moved forward along their route but they take up no space in the road network.Q2: no, vehicles are loaded in chunks of ~300 seconds and every bernoulli experiment is done only when the vehicle is loadedQ3: correctQ4: no, see Q2.Q5: yes. after insertion, the option device.rerouting.period determines the rerouting period and pre-period is no longer relevant.2018-09-12 22:52 GMT+02:00 Jane Cheung <cheung180515@xxxxxxxxx>:Jakob,Thanks for your detail _expression_ to my question.After your explanation, I would like to make sure that I really understand what you explained to me.for --teleportedQ1:Dose it mean thateven though we could not see the vehicle which is teleported ,this vehicle still driving in the network .Just like 'Stealth' ?for --device.rerouting.xxxx commands1) device.rerouting.deterministic is a boolean optionExample. You have set device.rerouting.probability to 0.33333333 and there are 10000 vehicles in your simulation.Without option device.rerouting.deterministic, you will have an expected number of 3333 vehicles with rerouting device but the actual number will vary according to a binomial distributionQ2:(Does it mean that it will do a sequence of 10000 dependent experiments with p=0.33333333 as soon as we start to simulate our simulation?,like Bernoulli experiment?)2) Every vehicle that is driving in the network will check for a faster route and adapt it's route if the current one isn't fastest once per minuteQ3:(Every vehicle that is loaded, but has not yet entered the network will not be controlled by this command)3) Every vehicle that is loaded but has not yet entered the network will l check for a faster route and adapt it's route if the current one isn't fastest once per minuteQ4:(all the vehicles will be loaded that the beginning of the simulation if they are pre-defined in the rou.xml)Q5:(the command could not be used to control these vehicles once they entered the network)Thanks at all!Best!JaneOn Wed, Sep 12, 2018 at 3:09 PM Jakob Erdmann <namdre.sumo@xxxxxxxxx> wrote:2018-09-12 21:57 GMT+02:00 Jane Cheung <cheung180515@xxxxxxxxx>:One more question,my understanding are listed below, are they correct?
--device.rerouting.deterministic false The devices are set deterministic using a fraction of 1000 (with the defined probability) --device.rerouting.period <STRING> 0 The period with which the vehicle shall be rerouted --device.rerouting.pre-period <STRING> 60 The rerouting period before insertion/depart for example:1)
--device.rerouting.deterministic=0.07 0.07 equals to 0.7(in a fraction of 100)2)For every vehicle, not
--device.rerouting.period =60
