Hi Mohsen,
I have the same work to do to make my simulation more realistic. I followed
roughly the following:
1- A colleague suggested a python program to generate random vehicle starts in
a xml file. I did not check whether it was consistent with my data but it seems
realistic.
Be careful with the syntax, it could not match SUMO definitions. For instance «
departlane » must be written like this : « departLane »… I changed it inthe
program below.
This is the code :
##############CODE#############################
import subprocess, sys, socket, time, struct, random
RUNS=180 #numbers of runs
N=1000 #number of cars per road
p1=0.5 #cars per second in road1
p2=0.5 #cars per second in road2
sumoexec= "sumo"
for j in range(RUNS):
routes = open("routes.rou.xml", "w")
print >> routes, '<routes>'
print >> routes, ' <vtypeDistribution id="random">'
print >> routes, ' <vtype vclass="vip" id="Car" accel="4.5"
decel="10" sigma="0.5" length="3" maxspeed="33" probability="0.9"/>'
print >> routes, ' <vtype vclass="delivery" id="Truck"
accel="1.8" decel="6" sigma="0.5" length="10" maxspeed="25" probability="0.1"/>'
print >> routes, ' </vtypeDistribution>'
print >> routes, ' <route id="route1" edges="road1 union1"
departLane="free" departpos="free"/>'
print >> routes, ' <route id="route2" edges="road2 union2"
departLane="free" departpos="free"/>'
lastcar1=0;
lastcar2=0;
for i in range(N):
nextcar1=lastcar1+random.expovariate(p1)
lastcar1=nextcar1
print >> routes, ' <vehicle route="route1" id="0_%i" type="random"
depart="%f" departLane="%i"/>' % (i,nextcar1,(i % 2))
nextcar2=lastcar2+random.expovariate(p2)
lastcar2=nextcar2
print >> routes, ' <vehicle route="route2" id="1_%i" type="random"
depart="%f" departLane="%i"/>'% (i,nextcar2, (i % 2))
print >> routes, "</routes>"
routes.close()
2- I made a python program to put departure times in the correct order.
Otherwise SUMO do not take into account.
#!/usr/bin python
import os
import re
import nltk
import collections
from collections import *
import codecs
import shutil
# //////////////////////////////////////////////////////////////
#
# **************** ordonner_routes poisson ********************************
#
# FAUT QUE LES TEMPS DEPART SE SUIVENT EN ORDRE CROISSANT DANS SUMO
# TIME DEPARTURE OF EACH VEHICULE MUST BE IN CHRONOLOGICAL ORDER
#
# //////////////////////////////////////////////////////////////
def List_route (INPUT_routes):
# fichier INITIAL routes.rou.xml
global route,routes,list_route
print "\n FICHIER INITIAL routes.rou.xml Temps depart ne sont pas en
ordre \n "
f_depart=open (INPUT_routes,"r")
lines = f_depart.readlines()
i=0
while i< len(lines):
print lines[i],
i=i+1
f_depart.close()
print "\n\n\n "
# partie du BAS
fo = open (INPUT_routes,"r,w")
fich_xml=fo.read()
p = re.compile('depart="[0-9]+.[0-9]+"')
route=p.findall(fich_xml)
routes=len(route)
fo.seek(0, 0)
fo.read
list_route=[]
i=0
while (i<routes):
n=route[i]
n_sub1 = re.sub('depart="', "", n)
n_sub2=re.sub('"',"",n_sub1)
list_route.append(n_sub2)
i=i+1
fo.close()
T2=map(float, list_route)
list_2=[]
T2_ordre=sorted (T2) # classe les temps de depart en ordre
i=0
while (i<len(T2)):
string_value = str(T2_ordre[i])
list_2.append(string_value)
i=i+1
List=[]
List = open(INPUT_routes).readlines()
List_bas=[] # liste du bas sera classe temps de depart
croissant
i=0
j=0
while i<len(List) :
j=0
for line in List:
if (i <len(list_2)):
temps_depart=list_2[i]
match = re.search(temps_depart, line)
new_line=re.sub('depart="[0-9]+.[0-9]+"','depart=\"' + temps_depart+'\"',line)
if match:
List_bas.append(new_line)
j=j+1
i=i+1
# partie du HAUT
fo=open(INPUT_routes, 'r')
lecture = fo.readlines()
N_lignes = len(lecture)
top=N_lignes-len(list_2)-1
fo.close()
List_top=[]
fo = open(INPUT_routes, 'r')
line = fo.readline()
List_top.append(line)
i=0
while i<top-1:
line = fo.readline()
List_top.append(line)
i=i+1
fo.close()
# FINAL au complet en ordre croissant ecran ET NOUVEAU FICHIER routes_2.rou.xml
# FINAL in chronological order
print "\n FICHIER FINAL routes_2.rou.xml Temps depart en ordre \n "
list_compl=List_top+List_bas
list_compl.append("</routes>")
long_list=len(list_compl)
i=0
while i< long_list:
print list_compl[i],
i=i+1
f = open("routes_2.rou.xml", 'w')
for item in list_compl:
f.write("%s" % item)
f.close()
print "\n\n "
3. Pay attention to the syntax because that can distort the running of the
program.
4- Michael Behrisch recommended me to correct certain errors in syntax and
replace the header to avoid syntax errors. It can detect syntax errors.
Michael recommandation :
Also please enable schema checking by using
the following header for your route file (instead of the <routes> line):
<routes xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance";;
xsi:noNamespaceSchemaLocation="http://sumo.dlr.de/xsd/routes_file.xsd";>;
This will point out misspelled or misused attributes automatically.
By bringing all these parties (and some adjustments), it seems to work
correctly.
I suggest also to see some tutorials in Excel (youtube) that can give you an
idea of poisson law.
Hope that help
François Vaudrin
Université Laval, CANADA
________________________________________
De : Jakob Erdmann [[email protected]]
Date d'envoi : 28 avril 2015 06:25
À : mohsen hs
Cc : [email protected]
Objet : Re: [sumo-user] Arrival rate, Distribution of vehicles within an hour
http://en.wikipedia.org/wiki/Poisson_distribution
http://en.wikipedia.org/wiki/Binomial_distribution
also see
http://sumo.dlr.de/trac.wsgi/browser/trunk/sumo/tools/randomTrips.py:323-328
for some code
2015-04-28 12:03 GMT+02:00 mohsen hs <[email protected]>:
> Hello there
> I am looking for a method to simulate the vehicles' distributions within
> an hour. As far as I know, OD matrix distributes vehicles within an hour
> uniformly. I was wondering if anybody could kindly provide me with some
> information, papers, or formulas to improve this distribution within an
> hour. The data that I have, counted the number of vehicle per hours.
> Since,I need a distribution model for simulation within an hour. Any Help
> is highly appreciated.
>
> Kind regardsMohsen
>
> ------------------------------------------------------------------------------
> One dashboard for servers and applications across Physical-Virtual-Cloud
> Widest out-of-the-box monitoring support with 50+ applications
> Performance metrics, stats and reports that give you Actionable Insights
> Deep dive visibility with transaction tracing using APM Insight.
> http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
> _______________________________________________
> sumo-user mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/sumo-user
>
------------------------------------------------------------------------------
One dashboard for servers and applications across Physical-Virtual-Cloud
Widest out-of-the-box monitoring support with 50+ applications
Performance metrics, stats and reports that give you Actionable Insights
Deep dive visibility with transaction tracing using APM Insight.
http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
_______________________________________________
sumo-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/sumo-user
------------------------------------------------------------------------------
One dashboard for servers and applications across Physical-Virtual-Cloud
Widest out-of-the-box monitoring support with 50+ applications
Performance metrics, stats and reports that give you Actionable Insights
Deep dive visibility with transaction tracing using APM Insight.
http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
_______________________________________________
sumo-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/sumo-user
