ArticlePDF Available

Car capacity near bus stops with mixed traffic derived by additive-conflict-flows procedure

Authors:

Abstract and Figures

To determinate car capacity at bus stops with mixed traffic, a new theoretical approach was developed on the basis of additive-conflict-flows procedure. The procedure was extended from homogeneous traffic flow to mixed traffic flow. The conflicts among cars, buses and bicycles near the stop can be described by the extended procedure. The procedure can be understood more easily than the theory of gap acceptance. Car capacity near the stop is the function of both bus stream and bicycle stream. The proposed model can also analyze the cases of pedestrian effects and limited priority of bicyclists. Numerical results show that the car capacity decreases with the increasing flow rates of other streams. In addition, pedestrian effects and bicyclist’s limited priority have negative effects on car capacity near bus stops with mixed traffic flow. Keywordsroad capacity–bus stop–additive-conflict-flows procedure–mixed traffic flow
Content may be subject to copyright.
Car
Car
Car
CarTravel
Tra vel
Travel
TravelTime
Tim e
Time
TimeEstimation
Esti mation
Estimation
Estimationnear
nea r
near
neara
a
a
aBus
Bus
Bu s
BusS
S
S
Stop
top
to p
topwith
with
w ith
withNon-motorized
Non -motorized
Non-motorized
Non-motorizedVehicles
Veh icles
Vehicles
Vehicles
Yang
Yang
Yang
YangXiaobao
Xia obao
Xiaobao
Xiaobao
SchoolofTrafficandTransportation,BeijingJiaotongUniversity
Beijing100044,China
Huan
Huan
Huan
HuanMei
Mei
M ei
Mei
SchoolofTrafficandTransportation,BeijingJiaotongUniversity
Beijing100044,China
E-mail:huanmei0125@sohu.com
Guo
Guo
Guo
GuoHongwei
Hon gwei
Hongwei
Hongwei
DepartmentofTransportationEngineering,BeijingInstituteofTechnology
Beijing100081,China
E-mail:07114190@bjtu.edu.cn
Gao
Gao
Gao
GaoLiang
Lia ng
Liang
Liang
DepartmentsofPhysics,Biology,andComputerScience,NortheasternUniversity
Boston,MA02115,USA
E-mail:lianggao2010@gmail.com
Abstract
Abstract
Abstract
Abstract
RealtimesystemforvehicletraveltimeandtrafficflowisanessentialpartofIntelligentTransportationSystems.
InmanyChinesecities,theinteractionsamongbuses,bicyclesandcarsbringdifficultytotraveltimepredictionand
trafficsafetymanagement.
T
heaimofthispaperistodevelopanewmodeltoestimatecartraveltimenearbus
stopsindevelopingcountriesbydataminingtechniquesandsurvivalanalysismethods.Thetraveltimedataunder
mixedtrafficconditionsarecollectedbyvideocamera.Fourinfluentialfactorsincludingcarvolume,non-
motorizedvolume,busdeparturevolumeandfreeratioofbusstoparechosenbyusingdataminingtechniques.A
proportionalhazard-baseddurationmodelisproposedtoanalyzethefactorsrelatedtocartraveltime.Theresults
indicatethatmixedtrafficflowimpactsthecartraveltimesignificantly.Inaddition,variousfactorscanmodifythe
traveltimedistributionindifferentdegreesandthemodelcanbeusedtoestimatethetraveltimeunderassumed
conditions.Itishopedtohelpimprovetheplanninganddesigningofproperfacilitieswithmixedtrafficflow.
Keywords
:
timeprediction;datamining;survivaldata;busstop;mixedtrafficflow;trafficsafety
Correspondingauthor.Tel.:+86-10-51687070.E-mailaddress:yangxb@bjtu.edu.cn.
International Journal of Computational Intelligence Systems, Vol. 4, No. 6 (December, 2011), 1350-1357
Received 9 July 2011
Accepted 29 November 2011
1.
1.
1.
1.Introduction
I ntroduction
Introduction
Introduction
Inrecentdecades,manystudieshavebeenconductedon
thetrafficsafetyonhighwaysandurbanintersections
1-3
.
Asadevelopingcountry,Chinahasherowntraffic
characteristics.
A
Mixofnon-motorizedandmotorized
vehiclesisanimportanttraffictypeinChina.Some
surveysshowthatthenon-motorizedvehicle,mainly
bicycleisoneofthemostwidelyusedtraffictoolsin
Chinesedailytravelactivity
4
.Althoughsome
researchersfocusedonthecar-bicycleconflictand
injuryaccidentsinvolvingcyclists
5-7
,littleinformation
hasbeenpublishedconcerningtheconflictamong
mixedtrafficstreamsnearbusstops.Meanwhile,with
thedevelopmentofnewtechnology,AdvancedTraffic
InformationSystemsarewidelyusedintraffic
managementandcontrol.Realtimesystemforvehicle
traveltimeandtrafficflowisanessentialpartof
AdvancedTrafficInformationSystems.Inmany
Chinesebusstops,theinteractionsamongbuses,
bicyclesandcarsbringdifficultytotraveltime
estimationandtrafficsafetymanagement.
T
herefore,it
isnecessarytodevelopanestimatingapproachfor
traveltimenearbusstopswithmixedtrafficflow.
Typically,therearethreetypesofbusstopsinurban
areas:thecurbsidestops,busbays,andbusboarders
8
.
Andcurbsidestopsarethemostcommonbusstopsin
manyChinesecities.Figure1showsthemixedtraffic
streamsatatypicalcurbsidestop.Therearetwolanes
ontheurbanroadway,thenon-motorizedlaneandthe
motorizedlane.
A
ndtherearethreetypesoftraffic
stream:bicycle,busandcar.Busstopsareusually
locatedonthenon-motorizedlane.Whenabusdwells
atthecurbsidestop,bicyclesmovetothemotorized
laneandgoroundthestoppedbus.Thus,thepresence
ofastoppedbuscreatesatemporaryconflictbetween
bicyclesandcars,increasingvehicletraveltimeand
trafficinjuryrisk.Similarphenomenamaybefoundin
otherAsiandevelopingcountries,forexample,India,
Malaysia,Vietnam,andCambodia.
Fig.1.Mixedtrafficstreamsamongbuses,carsandbicycles
nearacurbsidestop.
Onthebusstop,manystudieswereconductedon
theeffectsofbusstopsontrafficbehavior.Forexample,
FitzpatrickandNowlinusedcomputersimulationto
determinehowbusstopdesigninfluencestraffic
operationsaroundabusstop
9
.LevinsonandJacques
usedfieldstudiesandsimulationanalysestovalidate,
update,andextendexistingbusstopandberthcapacity
procedures
10
.Wongetal.developedasimulation
modelforestimatingthedelayatasignal-controlled
intersectionwithabusstopupstream
11
.Tangetal.used
amacrodynamicmodeltostudytheeffectsofbusstop
ontrafficflow
12
.
U
nfortunately,littleinformationhas
beenpublishedconcerningbusstopswithnon-motor
vehicles.
Duetothespecialfeaturesofmixedtraffic,the
applicationofexistingtrafficmodelsforbusstopshas
notproducedacleareffectinChinesetraffic
managementandcontrol.Inrecentyears,some
researchershaverealizedthisparticularity,and
correlativeworkisbeingdone,butitstillhasalong
waytogo.Forexample,KoshyandArasanused
simulationtechniquetostudytheimpactofbusstop
typeonthespeedsofothervehiclesunder
heterogeneousconditions
8
.Zhaoetal.developeda
combinedcellularautomatamodelfortheinvestigation
ofroadcapacitynearabusstopwithnon-motorized
vehicles
13
.Yangetal.establishedtwomodelsforcar
capacitynearthecurbsidestopwithbicyclesbasedon
gapacceptancetheoryandtheadditive-conflict-flows
procedure,respectively
14-15
.
T
heyanalyzedtheeffects
ofbusstoponthevehiclespeedandcapacityunder
mixedtrafficconditions;however,theydidnotstudy
theeffectsofbusstoponthevehicletraveltime.
In
thispaper,weproposeahazard-basedduration
approachtoanalyzecartraveltimedistributionnearbus
stopsundermixedtrafficconditions.Thehazard-based
durationmodelshavebeenusedextensivelyin
biometricsandreliabilityengineeringfordecades
16
.
Durationmodelscanbeusedtodeterminecausalityin
durationdataandtheyarealsousefultoolsinthefield
oftransportation
17-19
.Thesemodelsrepresentatypeof
analyticalmethodstodescribethedurationofacertain
stateandhowvariousfactorshaveaffectedtheduration.
Itistheveryreasonwhythedurationmethodischosen
toanalyzethetraveltimedistributionundermixed
trafficconditions.Theempiricaldataaremodeledby
proportionalhazardfunction.Thefactorsareconsidered
Published by Atlantis Press
Copyright: the authors
1351
YangXiaobaoetal.
asinfluencevariablesincludingcarvolume,non-
motorizedvehiclevolume,busdeparturevolume,free
ratioofbusstop,andsoon.Withabovemethods,the
distributionofcartraveltimeundervariousconditions
iscalculatedandtheinfluenceoftheselectedvariables
isquantified.
2.
2.
2.
2.Methodology
M ethodology
Methodology
Methodology
2.1.
2.1.
2.1.
2.1.
Hazard-
Hazard-
Hazard-
Hazard-b
b
b
based
ased
ased
asedd
d
d
duration
uration
uration
urationm
m
m
model
odel
odel
odel
Let
T
beanonnegativerandomvariablerepresenting
thecartraveltimeinatestroadsection.Let
()
ft
denotetheprobabilitydensityfunctionof
T
andletthe
cumulativedistributionfunctionbe
0
()Pr()()
t
FtTtfudu
==
(1)
Let
()
St
denotetheprobabilitythatthetravel
durationdoesnotendpriorto
t
,yield
()Pr()()
t
StTtfudu
=>=
(2)
()
St
isusuallycalled“survivorprobabilityor
“enduranceprobability”inthedurationliterature.Inthis
paper,
we
define
()
St
ascontinuanceprobabilityin
ordertorepresenttheprobabilitythatacartravels
longerthan
t
(thetraveldurationstill“continues”at
t
).
Inthehazard-baseddurationapproach,
T
canbe
characterizedbyahazardfunction,
()
t λ
.Itrepresents
theinstantaneousprobabilitythatthetravelduration
willendinaninfinitesimallysmalltimeperiod,
t
,
aftertime
t
,giventhatthedurationhasnotendeduntil
time
t
.Themathematicaldefinitionforthehazard
functionintermsofprobabilitiesis
0
Pr()
()lim
t
tTttTt
t
t
λ
<+
=
(3)
T
hehazardfunctiongivestheconditionalfailurerate.
Inthisstudy,theconditionalfailurerateisthe
conditionalpassratethatcarspassthroughabusstop.
Thehazardfunctionistheinstantaneousrateatwhicha
carpassesthroughabusstopinaninfinitesimallysmall
timeperiod,
t
,aftertime
t
,giventhatthecarhasnot
passedthestopuntiltime
t
.
Theresultinthehazardfunctionishazardrateor
hazard.Specifically,
()
tt λ
istheapproximate
probabilityofthedurationterminatingin
[
)
,
ttt
+
,
givencontinuanceupto
t
.Thefunctions
()
Ft
,
()
St
,
()
ft
and
()
t λ
givemathematicallyequivalent
specificationsofthedistributionof
T
.Sothehazard
functioncanalsobedefinedintermsof
()
Ft
,
()
St
,
and
()
ft
,yield
()()ln()
()
1()()
λ
===
ftftd St
t
FtStdt
(4)
IntegratingEq.(4)fromzeroto
t
andusing
(0)1
S
=
,
yield
0
()exp(())
t
Studu λ
=
(5)
Notethatcartraveltimenearabusstopis
influencedbyvariousfactors.
A
primaryobjectiveof
thispaperistoaccommodatetheeffectsofthese
influentialfactors.Theinfluentialfactorscanbedefined
asavectorofexplanatoryvariables,
x
x
x
x
.Thenthe
proportionalhazard(PH)formisintroduced,which
specifiestheeffectsofexplanatoryvariablestobe
multiplicativeonahazardfunction,yield
0
()()(,)
ttg λλ
=x
x
x
xα
α
α
α
(6)
where
0
()
t λ
isthebaselinehazardfunctionrepresenting
thehazardwhentheeffectsofexplanatoryvariablesare
neglected[i.e.
(,)1
g
= x
x
x
xα
α
α
α
],
()
g
i
isaknownfunctionto
representtheeffectsofexplanatoryvariables,
α
α
α
α
isa
vectorofparametersforx
x
x
x.Inthispaper,atypical
specificationwith
(,)exp()
g
= x
x
x
xααx
ααx
ααx
ααx
,whichwas
proposedbyCox
20
,isused.Thisspecificationis
convenientsinceitguaranteesthepositivityofthe
hazardfunctionwithoutplacingconstraintsonthesigns
oftheelementsof
α
α
α
α
.TheCoxproportionalhazard
modelis
0
()()exp()
tt λλ
=αx
α x
αx
αx
(7)
ThecontinuanceprobabilityfunctioncombiningEq.
(5)andEq.(7)canbewrittenas
exp()
0
()[()]
StSt
=
αx
α x
αx
αx
(8)
where
0
()
St
isthebaselinecontinuanceprobability
function.
2.2.
2.2.
2.2.
2.2.
Model
Model
Model
Modele
e
e
estimation
stimation
stimation
stimation
ThemodelinEq.(8)hastwocomponents,
α
α
α
α
and
0
()
t λ
.
Cox
20
introducedaningeniouswayofestimating
α
α
α
α
;
thisisnowknownasthepartiallikelihoodmethod.
Becauseofitssimplicityandusefulness,methodology
relatedtothisapproachisadopted.Supposethata
randomsampleconsistsof
k
distinctobservedduration
data,
(1)(2)()
k
ttt
<<<
.Let
()
i
x
x
x
x
bethevariable
associatedwiththesampleobservedat
()
i
t
.Let
()
()
i
t
R
R
R
R
denotetherisksetat
()
i
t
sinceitconsistsofall
individualswhosedurationsareatleast
()
i
t
.Thelog-
partiallikelihoodfunctionforestimating
α
α
α
α
is
Published by Atlantis Press
Copyright: the authors
1352
CarTravelTimeEstimation
()
()
1()
()logexp()
i
k
il
ilt
LL
=
=
R
R
R
R
ααxαx
ααxαx
ααxαx
ααxαx
(9)
Theestimationof
0
()
t λ
canrefertoRef.18.
Theoverallgoodness-of-fitofthemodelestimation
isdeterminedbythelikelihoodratio(LR)statistics,
whichisspecifiedas
0
ˆ
2[()()]
L
XLL
= αα
α α
αα
αα
(10)
where
0
()
L
α
α
α
α
isthelog-likelihoodfornullmodelwith
alltheregressioncoefficientsaresetaszeroand
ˆ
()
L
α
α
α
α
isthelog-likelihoodatconvergencewith
k
regression
coefficients.
3.
3.
3.
3.Survey
Survey
Survey
Surveyand
an d
and
andData
D ata
Data
Data
3.1.
3.1.
3.1.
3.1.
Site
Site
Site
Sites
s
s
survey
urvey
urvey
urveyd
d
d
design
esign
esign
esign
Cartraveltimeistheactuallyobservedtimeforacarto
passabusstopundermixedtrafficconditions.To
recordthetraveltimedurations,twovideoswereplaced
theupstreamsectionandthedownstreamsectionofthe
busstop,respectively.Everycarpassingthecurbside
stopwasobservedasadatacollectionunit.Thetravel
timedurationwasfromthetimethatacararrivesatthe
upstreamsectiontothetimethatitleavesthe
downstreamsection.Thestudyalsoexploredtheeffects
ofmixedtrafficflowonthetraveltime.Sothemixed
trafficcharacteristicsat1minintervalwereextracted
fromthevideos,suchastrafficvolumesofvarious
streamsandtheaveragedwelltimeofbusstreamat
everyminute.Meanwhile,thecartraveltimeofselected
sampleassociatedwiththemixedtrafficcharacteristics
wasmatchedaccordingtotheintervalthesample
belongedto.
Theobservedroadwaynearthestopscontainsa
non-motorizedlaneandamotorizedlane.Thesite
surveywasconductedattheselectedcurbsidestopon
XueyuanroadinBeijing,China.Datawerecollectedin
Januaryof2008.Thesurveyperiodsincludedpeakhour
andoff-peakhour.Thelengthofthetestsectionwas
67.5mandthewidthofthesectionswas7.5m.
In
addition,thetestsection
w
asnotinfluencedbythe
trafficcontrol.Thesurveywasconductedingood
weather.Basedonthetraveltimestudy,531data
sampleswereobserved.
3.2.
3.2.
3.2.
3.2.
Variable
Variable
Variable
Variables
s
s
specifications
pecifications
pecifications
pecifications
Theexplanatoryvariablesforinclusioninthemodel
werechosenbyusingdataminingtechniquesonthe
basisofpreviousresearchandintuitivearguments
regardingtheeffectofmixedtrafficflow.Inurban
traffic,traveltimeisdeterminedbydrivingbehavior,
trafficconditions,geometricdesign,andsoon.Inthis
paper,weconcernthecharacteristicsoftraveltime
undertheinfluenceofmixedtrafficflownearbusstops.
Therefore,thevariablespecificationsusedinthe
durationmodelshouldberelatedtomixedtrafficflow.
Consideringthefeasibilityofdataacquisition,four
explanatoryvariablesincludingcarvolume(
X
c
),non-
motorizedvehiclevolume(
X
n
),busdeparturevolume
(
X
b
),andfreeratioofbus stop(
X
f
)arechosen.
T
hefree
ratioofbusstopreflectstheprobabilityofnobusatthe
stopateveryminute.Althoughthevolumeofthe
comingbushasanimpactonthecartraveltime,thereis
asignificantlystatisticalcorrelationbetweenthevolume
ofthecomingbusandthevolumeofthedeparturebus.
A
dditionally,theconflictbetweenthedeparturebusand
thepassingcarmayoccur.Thedeparturebushasa
greaterimpactonthecartraveltimethanthecoming
bus.Thus,notthevolumeofthecomingbusbutthe
volumeofthedeparturebusischosen.Analogously,
thereisasignificantlystatisticalcorrelationbetweenthe
dwelltimeofstoppedbusandfreeratioofbusstop.
Becausethelatterhasamoredirectinfluenceonthecar
traveltimethantheformer,thedwelltimeofstopped
busisnotchosen.
4.
4.
4.
4.Empirical
Empirical
Empirical
EmpiricalResults
R esults
Results
Results
4.1.
4.1.
4.1.
4.1.
Estimated
Estimated
Estimated
Estimatedr
r
r
results
esults
esults
esults
Theestimationofthedurationmodelforcartraveltime
isshowninTable1.TheLRstatisticoftheestimated
modelclearlyindicatestheoverallgoodness-of-fit(the
LRstatisticis5,471.4,whichisgreaterthanthechi-
squaredstatisticwith4degreesoffreedomatany
reasonablelevelofsignificance).Ontheotherhand,the
statisticalsignificanceofeachvariablealsoindicatesthe
significantpresenceofvariablesinthedurationofthe
traveltime.Fromtheresults,alloftheincluded
variablesarestatisticallysignificantatthe0.02levelof
significance.
Published by Atlantis Press
Copyright: the authors
1353
YangXiaobaoetal.
Table1.Modelestimation.
VariableCoefficient
Standard
Error
Z-statisticPValue
X
c
-0.0410.018-2.3620.018
X
n
-0.0300.009-3.2110.001
X
b
-0.4310.054-7.9160.000
X
f
1.0590.1756.0490.000
Fig.2.Observedcontinuanceprobabilityandestimated
continuanceprobability.
Figure2showstheestimatedcontinuance
probabilitybytheproposedmodelandtheobserved
continuanceprobability.Thecurveofestimated
distributionismonotonedecreasing.Themedianofthe
distributionis14.4s,indicatingthatoverahalfofthe
observedvehiclescanpassthetestsectionwithin14.4s.
The25%quantileofthedistributionis16.7s,indicating
thatabout25%oftheobservedvehiclescannotpassthe
testsectionwithin16.7s.Theobservedresultsshowthe
sameoverallshapeastheestimatedresults.Whilethe
generalshapeisthesamebetweenthetworesults,the
observedcontinuanceprobabilityisslightlylargerthan
theestimatedresults.
T
hisdiscrepancyiscausedbythe
differentsourcesoftwocontinuanceprobabilities.In
Figure2,theobserveddistributionisobtainedfromthe
cumulateddistributionofobservedtraveltime.Andthe
estimateddistributioniscalculatedbythemodelwith
averagevariables.
T
herefore,theobservedresults
indicatethetraveltimeunderthespecificconditionfor
individualsample,whiletheestimatedresultsindicate
theaverageconditionthatisrelatedwithdifferent
influentialfactorsandallsamples.
T
heestimated
continuanceprobabilityshowninFigure2canreflect
thecharacteristicsoftraveltimewhichhasanaverage
valueforeveryvariable.Anychangeofthetraffic
conditionscouldinfluencethecartraveltime.The
effectsofvariablesarediscussedinthenextsection.
4.2.
4.2.
4.2.
4.2.
Effects
Effects
Effects
Effectsof
of
of
ofExplanatory
Explanatory
Explanatory
ExplanatoryVariables
Variables
Variables
Variables
AccordingtoEq.(7),theeffectsoftheexplanatory
variablescanbeinterpretedbythesignsofthe
coefficientsinaratherstraightforwardfashion.Ifthe
coefficientisnegative,itimpliesthatanincreaseinthe
correspondingvariabledecreasesthehazardrate,or
equivalently,increasesthecartravelduration.With
regardtothemagnitudeofthevariableeffects,whena
variablechangesbyoneunit,thehazardwouldchange
by
[exp()1]100%
i
β
×
.AsshowninTable1,the
variable
X
f
indicatespositiveeffectontravelduration.
Othervariablesshowanegativeeffectthatthe
increasingvariablescouldincreasethecartraveltime.
Toassesstheeffectsoftheincludedexplanatory
variablesonthetravelduration,afunctionofhazard
ratio(HR)canbeobtainedbydividingbothsidesofEq.
(7)by
0
()
t λ
,yield
1122
0
()
exp()exp()
()
iirri
t
xxx
t
λ
ααα
λ
==+++ αx
α x
αx
αx
(11)
where
ri
x
isthe
r
thvariableforthe
i
thvehicle,
α
r
is
thecorrespondingcoefficient.TheHRcanrepresentthe
multiplerelationsbetweenthehazardundertheeffects
ofvariablesandthehazardwhenallvariablesare
ignored(
0 = x
x
x
x
).
Thevariablesinthedenominatoroftheleftsideof
Eq.(11)arestandardizedaboutthemeanandyield
111222
()
exp[()()()]
()
iirrir
t
xxxxxx
t
λ
ααα
λ
=+++
(12)
where
()
t λ
isthehazardwiththeaveragevariables,
r
x
istheaverageofthe
r
thvariableforallsample.
Eq.(12)istherelativehazardratio(RHR,itisalso
calledtherelativehazardindex).Itrepresentstheratio
ofthehazardforavehiclewithagivensetofvariables
tothehazardforavehiclewhichhasanaveragevalue
foreveryvariable.IftheRHRismorethanone,it
meanstheeffectsofthevariablescanincreasethe
hazardandsothevariablesarefavorable.Thatistosay,
thetraveltimeinsuchfavorableconditionislessthan
theaveragelevelofthesurveysample.Onthecontrary,
theunfavorablevariablescorrespondtoalowhazard.
Therefore,thevehiclesintheunfavorablecondition
wouldhavelongertraveltimethanthoseinthe
favorablecondition.Inordertomakeaquantitative
Published by Atlantis Press
Copyright: the authors
1354
CarTravelTimeEstimation
analysisoftheeffectsofthemixedtrafficflow,wecan
assumethatavariableisinthefavorableorunfavorable
conditionandothervariablestaketheiraveragevalues.
ThentheHRorRHRforeachvariablecanbecalculated.
Inaddition,theRHRcanbeusedtodescribethe
multipleofthehazardwhentheobservedvehiclesarein
favorableconditionorunfavorableconditioncompared
withtheaveragecondition.TheHRcanbeusedto
describethemultipleofthehazardwhentheobserved
vehiclesareinfavorableconditioncomparedwiththe
unfavorablecondition.AccordingtotheRHRsandHRs
inspecificconditions,theinfluenceofmixedtraffic
flowonthetraveltimecanbeanalyzedquantitatively.
ThespecificconditionsandcorrespondingHRsand
RHRsareshowninTable2.
Table2.Analysisofvariablesrelatedtotraveltime.
VariableMean
VariablesValue
RelativeHazard
Ratio
Hazard
Ratio
UnfavorableFavorable
Low
Hazard
High
Hazard
X
c
11.8430.0010.000.481.082.25
X
n
13.9930.0010.000.621.131.82
X
b
1.843.001.000.611.442.36
X
f
0.500.100.900.661.532.32
Toillustratetheeffectsoftheselectedvariables,the
RHRsforfourvariables(
X
c
,
X
n
,
X
b
and
X
f
)areshownin
Figure3.AsshowninFigure3(a),thevariable
X
c
(car
volume)indicatesanegativeeffectonthehazard.This
reasonisthattheincreasingcarvolumecouldincrease
theprobabilityofcarqueueandthendelaythecartravel
time.TheRHRis2.25forthespecificconditions.It
meansthatthecarpassratewith10carsis2.25times
thatforacarwith30cars,whenothervariablestake
theiraveragevalues.Comparedtothefiledsurvey,the
averagetraveltimeofthevehicleswithcarvolumeless
than11.84is13.85s;whiletheaveragetraveltimewith
carvolumemorethan11.84is15.67s.
Thevariable
X
n
(non-motorizedvehiclevolume)
indicatesanegativeeffectonthehazard,whichis
showninFigure3(b).Whenabusdwellsatthestop,
non-motorizedvehiclesmaychangetothemotorized
lane.Thus,theincreasingvolumefornon-motorized
vehiclescouldcausethelargerprobabilityofthe
conflictbetweencarsandnon-motorizedvehicles,
whichfinallyleadtotheincreaseofcartraveltime.The
RHRis1.82forthespecificconditions;thatis,thepass
Fig.3.Relativehazardindexesforfourvariables
rateforacarwith10non-motorvehiclesis1.82times
thatforacarwith30non-motorvehicles.
Thevolumeofbusdeparturealsoshowsanegative
effectontraveltime.Thisisthereasonthatthecar-bus
conflicttakesplacewhenabusdepartsfromthestopto
themotorizedlane.Thecar-busconflictcoulddelaythe
cartraveltime.TheRHRis2.36forthespecific
conditions.Itmeansthatthecarpassratewithonlyone
departurebusis2.36timesthatwith3departurebuses
[seeFigure3(c)].
Theeffectoffreeratioofbusstop(
X
f
)indicatesthat
theincreasingfreeratioofbusstopcanincreasethe
hazard,ordecreasethecontinuanceprobability[see
Figure3(d)].
It
meanstheprobabilitythatthevehicles
cantransversethebusstoplongerthantime
t
would
decrease.Thiscanbeexplainedbytheeffectofstopped
bus.Ifnobusdwellatthestop,thereisnotheconflict
betweencarsandnon-motorizedvehicles,andthenon-
motorizedvehicleshavenoimpactonthecartravel
time.Otherwise,oneormorestoppedbusesdwellatthe
stop,theconflictbetweencarsandnon-motorized
vehiclesmayleadtotheincreaseofthecartraveltime.
Thus,theincreasingfreeratioofbusstopmayincrease
theconflictbetweencarsandnon-motorizedvehicles
andfinallyleadtotheincreaseofthecartraveltime.
FromtheresultsinTable2,thehighfreeratioofbus
stop(90%)is2.32timesaslongasthelowfreeratio
(10%)tomakethevehicleshavelongertraveltimes.
Accordingtotheobserveddata,theaveragetraveltime
ofthevehicleswithfreeratioofbusstoplessthan50%
is17.35s;whiletheaveragetraveltimewithfreeratio
morethan50%is12.57s.Theobservedresultsalso
verifytheestimatedresults.
Published by Atlantis Press
Copyright: the authors
1355
YangXiaobaoetal.
5.
5.
5.
5.Model
M odel
Model
ModelApplication
A pplication
Application
Application
Theestimatedmodelcanbeusedtopredictthe
distributionofcartraveltimeunderdifferentconditions.
Byusinganassumedvariableunderaspecified
condition,anewdistributionofthecontinuance
probabilitycanbecalculatedwhileothervariablesareat
theirmeanvaluesrespectively.Inthispaper,thenon-
motorizedvehiclevolumeistakenasanexampleto
presentthemodelapplication.Ifthenon-motorized
vehiclevolumerangesfrom5to35veh/min,thenew
distributionsofthecontinuanceprobabilityareshownin
Figure4.Thedifferencesbetweenthecurvesindicate
thestrongeffectofthenon-motorizedvehiclevolume.
Whilethenon-motorizedvehiclevolumeincreasesfrom
5to35veh/min,themediansare13.5s,14.5s,and16.4
s;the25%quartilesare15.3s,17.6s,and20.7s.Here,
the25%quantileisdefinedastheapproximatetravel
time.Thetraveltimewouldhaveanincreaseof17.6%
ifthenon-motorizedvehiclevolumeincreasesfrom20
to35veh/min.Othervariablesestimatedbytheduration
modelcanbeusedtopredictthedistributionofthe
continuanceprobabilityinthesameway.
Fig.4.Distributionsofthecontinuanceprobabilitywith
differentnon-motorizedvehiclevolumes
Thetraveltimeistypicaldurationdatathatisthe
concernofthedurationmodel.Mostimportantly,the
hazard-baseddurationmethodologycancapturethe
effectsofmixedtrafficflownearbusstops.Therefore,
thehazard-baseddurationapproachcouldbehelpfulin
thetraveltimeestimationnearbusstopswithmixed
trafficflow.Beforetheapplicationstoothersites,the
modelshouldbeestimatedusingthespecifiedfielddata.
Additionally,theexplanatoryvariablescanbechosen
flexiblyaccordingtotheresearchaimandthetraffic
reality.
6.
6.
6.
6.Conclusions
C onclusions
Conclusions
Conclusions
RealtimesystemisnecessaryforAdvancedTraffic
ManagementSystems.Thispaperappliesahazard-
baseddurationmodeltoestimatecartraveltimenear
busstopsundermixedtrafficconditions.Theconflict
amongbuses,bicyclesandcarsarediscussed.Thetravel
timestudiesareconductedinthetestedbusstopwith
non-motorizedvehicles.Meanwhile,mixedtrafficflow
andtheireffectsonthecartraveltimearerecorded.The
methodologyusesaframeworkofproportionhazard
thatallowsdifferentindividualstohavevarioustravel
timesaccordingtothetrafficconditions.
Thepaperprovidesseveralimportantinsightsinto
thedeterminantsofthecartraveltimedistributionnear
busstopsindevelopingcountries.Firstly,theresults
indicatethatcartraveltimeisaffectedbyvarious
relatedfactorsnearabusstop.Suchinfluencecanbe
reflectedbythedistributionofcartraveltime.Any
changeofaninfluentialfactorcouldchangethetravel
timedistribution.Secondly,thefreeratioofbusstop
showsanegativeeffectonthecartraveltime,whilethe
carvolume,thenon-motorizedvehiclevolumeandthe
busdeparturevolumeshowpositiveeffectsonthecar
traveltime.Additionally,fromthemethodological
standpoint,thisstudyhasprovidedtheempirical
evidencethathazard-baseddurationapproachis
appropriateforthetraveltimeanalysisundermixed
trafficconditions.Thedistributionoftraveltime
estimatedbythemodelwouldgiveaquantitative
analysisoftheinfluenceofmixedtrafficflow.Finally,
theinfluentialfactorsrelatedmixedtraffic
characteristicsshouldbegivenfullconsiderationinthe
planninganddesigningofbusstops.
In
termsofthefuturework,researchwithmore
datasetsisrequired.Alsootherinfluentialfactorsshould
beconsidered,suchasthenumberofpassengerloads,
andthetypeofbusstops.Inaddition,i tisnecessaryto
studyvehiclespeeddistributionnearabusstopwith
mixedtrafficflow.Itishopedthatthesefindingsmay
haveabetterunderstandingofbusstopsandhelpto
plananddesigntrafficfacilitiesindevelopingcountries.
Acknowledgements
Acknowledgements
Acknowledgements
Acknowledgements
ThisworkissupportedbyNationalNaturalScience
FoundationofChinaunderGrant70901005,71131001,
SpecializedResearchFundfortheDoctoralProgramof
HigherEducationunderGrant20090009120015,and
Published by Atlantis Press
Copyright: the authors
1356
CarTravelTimeEstimation
FundamentalResearchFundsfortheCentral
UniversitiesunderGrant2011JBM055.
References
References
References
References
1.W.H.Wang,Q.Cao,K.
i
Ikeuchi,andH.Bubb,Reliability
andsafetyanalysismethodologyforidentificationof
drivers'erroneousactions,
Int.J.Automot.Techn.
11
11
11
11(6),
(2010)873-881.
2.W.H.Wang,F.G.Hou,H.C.Tan,andH.Bubb,A
frameworkforfunctionallocationinintelligentdriver
interfacedesignforcomfortandsafety,
Int.J.Comput.
Int.Sys
,3
3
3
3(5),(2010)531-541.
3.W.H.Wang,
Y
.Mao,J.Jing,X.Wang,H.W.Guo,X.M.
Ren,andI.Katsushi,Driver’svariousinformation
processandmulti-ruleddecision-makingmechanism:a
fundamentalofintelligentdrivingshapingmodel,
Int.Nt.
J.Comput.Int.Sys
.4
4
4
4(3),(2011)297-305.
4.D.H.Wang,T.J.Feng,C.Y.Liang,et
al.
Researchon
bicycleconversionfactors,
TransResA
,42
4 2
42
42(8),(2008)
1129-1139.
5.I.Walker,Driversovertakingbicyclists:objectivedataon
theeffectsofridingposition,helmetuse,vehicletypeand
apparentgender,
AccidentAnalysisandPrevention,
39
39
39
39,
(2007)417-425.
6.J.M.Wood,P.F.Lacherez,R.P.Marszalek,andM.J.
King,Driverscyclyists’experiencesofsharingtheroad:
incidents,attitudesandperceptionsofvisibility,
Accident
AnalysisandPrevention,
41
41
41
41,(2009)772-776.
7.S.Daniels,T.Brijs,E.Nuyts,andG.Wets,Injurycrashes
withbicyclistsatroundabouts:influenceofsomelocation
characteristicsandthedesignofcyclefacilities,
Journal
ofSafetyResearch,
40
40
40
40(2),(2009)141-148.
8.R.Z.Koshy,andV.A.Arasan,Influenceofbusstopson
flowcharacteristicsofmixedtraffic,
J.Transp.Eng
.
131
131
131
131(8)(2005)640-643.
9.K.Fitzpatrick,andR.L.Nowlin,Effectsofbusstopdesign
onsuburbanarterialoperations,
Transport.Res.Rec.
1571
1571
1571
1571(1997)31-41.
10.H.S.Levinson,K.R.S.Jacques,Buslanecapacity
revisited,
Transport.Res.Rec.
1
1
1
1618
6 18
618
618(1998)189-199.
11.S.C.Wong,H.Yang,W.S.
Y
eung,etal.Delayat
signal-controlledintersectionwithbusstopupstream,
J.
Transp.Eng
.124
1 24
124
124(3)(1998)229-234
12.T.Q.Tang,
Y
.Li,andH.J.Huang,Theeffectsof
busstopontrafficflow,
Int.J.Mod.Phys.C
,20
2 0
20
20(6)
(2009)941-952.
13.X.M.Zhao,B.Jia,Z.Y.Gao,andK.P.Li,Traffic
interactionsbetweenmotorizedvehiclesand
nonmotorizedvehiclesnearbusstop,Journalof
TransportationEngineering,
J.Transp.Eng
.135
1 35
135
135(11)
(2009)893-905.
14.X.B.Yang,Z.Y.Gao,X.M.
Z
hao,andB.F.Si,
Roadcapacityatbusstopswithmixedtrafficflowin
China,
Transport.Res.Rec.
211
211
211
2111
1
1
1(2009)18-23.
15.X.B.Yang,Z.
Y.
Gao,B.F.Si,andL.Gao.Car
capacitynearbusstopswithmixedtrafficderivedby
additive-conflict-flowsprocedure,
Sci.ChinaTechnol.
Sci.
54
54
54
54(3)(2011)737-740.
16.N.M.Kiefer,Economicdurationdataandhazard
functions,
JournalofEconomicLiterature
,26
2 6
26
26(2)(1988)
646-679.
17.D.A.Hensher,andF.L.Mannering,Hazard-based
durationmodelsandtheirapplicationtotransport
analysis,
Transp.Rev.
14
14
14
14(1)(1994)63-82.
18.C.R.Bhat,Durationmodeling.Handbookof
transportmodelling,in
TransportModelingAnnuals
(ElsevierScience,2000),pp.91–111.
19.H.W.Guo,Z.Y.Gao,X.B.Yang,andX.B.Jiang,
Modelingpedestrianviolationbehavioratsignalized
crosswalksinChina:ahazards-baseddurationapproach,
TrafficInj.Prev.
12
12
12
12(1)(2011)96-103.
20.D.R.Cox,Regressionmodelsandlifetables,
J.Roy.
Stat.Soc.B
,34
3 4
34
34(2)(1972)187-220.
Published by Atlantis Press
Copyright: the authors
1357
... Koshy and Arasan used simulation technique to study the impact of bus stop type on the speeds of other vehicles under heterogeneous conditions [9]. Yang et al. established car capacity models near a curbside stop with bicycles based on gap acceptance theory and conflict technique [12,13]. However, little information was found in the literature on delay time near bus stops with mixed traffic flow. ...
... Bicycles in the nonmotorized lane would force the subsequent car in the motorized lane to slow down for the lane-changing execution. Field observations indicate that this cooperative lane changing and priority-sharing behavior is prevalent between bicycles and cars near bus stops [13]. As the acceptable gap of bicycles is approximate to the follow-up time of successive cars, the bicycle-car conflict near a bus top is similar to the conflict at merges under low speed or high flow conditions. ...
... The car delay caused by the car-bicycle conflict is equal to the expected waiting time in the queue system for mixed streams between cars and bicycles, ,B . As the mean service time for car stream at point B is larger than that for bicycle stream, ,B can be obtained by (13). Thus, the car delay at the point B can be given as ...
Article
Full-text available
This paper proposes a model for estimating car delays at bus stops under mixed traffic using probability theory and queuing theory. The roadway is divided to serve motorized and nonmotorized traffic streams. Bus stops are located on the nonmotorized lanes. When buses dwell at the stop, they block the bicycles. Thus, two conflict points between car stream and other traffic stream are identified. The first conflict point occurs as bicycles merge to the motorized lane to avoid waiting behind the stopping buses. The second occurs as buses merge back to the motorized lane. The average car delay is estimated as the sum of the average delay at these two conflict points and the delay resulting from following the slower bicycles that merged into the motorized lane. Data are collected to calibrate and validate the developed model from one site in Beijing. The sensitivity of car delay to various operation conditions is examined. The results show that both bus stream and bicycle stream have significant effects on car delay. At bus volumes above 200 vehicles per hour, the curbside stop design is not appropriate because of the long car delays. It can be replaced by the bus bay design.
... Tirachini et al. [11] evaluated the impact of the passenger crowding at the bus stops on the operations and travel time of buses. Some other studies focused on evaluating the conflicts between different road users near the bus stop areas [14][15][16][17][18]. For example, Zhao et al. [16] evaluated the traffic interactions between the motorized and nonmotorized vehicles near a bus stop. ...
Article
Full-text available
On urban streets with bus stops, bus arrivals can disrupt traffic flows in the neighboring areas. Different stop designs have distinct influences on the road users. This study aims to evaluate how different types of bus stops affect the operations of vehicles, bicycles, and buses that pass by. Four types of stops that differ in geometric layout are examined. They are termed the shared bike/bus (Type 1), separated shared bike/bus (Type 2), vehicle/bus with inboard bike lane (Type 3), and bus bay with inboard bike lane (Type 4). Data are collected from eight sites in two cities of China. Results of data analysis show that different bus stop designs have quite different impacts on the neighboring traffic flows. More specifically, Type 3 stops create the least bicycle delay but the largest vehicle delay. Type 4 stops have the least impact on bicycle and vehicle operations, but occupy the most road space. Traffic operations are less affected by Type 1 stops than by Type 2 stops. Policy suggestions are discussed regarding the optimal design of bus stops that minimizes the total vehicle delay of all modes.
... Levinson and Jacques used field studies and simulation analyses to validate, update, and extend existing bus stop and berth capacity procedures 10 13 . Yang et al. established two models for car capacity near the curbside stop with bicycles based on gap acceptance theory and the additive-conflict-flows procedure, respectively [14][15] . They analyzed the effects of bus stop on the vehicle speed and capacity under mixed traffic conditions; however, they did not study the effects of bus stop on the vehicle travel time. ...
Article
Full-text available
Real time system for vehicle travel time and traffic flow is an essential part of Intelligent Transportation Systems. In many Chinese cities, the interactions among buses, bicycles and cars bring difficulty to travel time prediction and traffic safety management. The aim of this paper is to develop a new model to estimate car travel time near bus stops in developing countries by data mining techniques and survival analysis methods. The travel time data under mixed traffic conditions are collected by video camera. Four influential factors including car volume, nonmotorized volume, bus departure volume and free ratio of bus stop are chosen by using data mining techniques. A proportional hazard-based duration model is proposed to analyze the factors related to car travel time. The results indicate that mixed traffic flow impacts the car travel time significantly. In addition, various factors can modify the travel time distribution in different degrees and the model can be used to estimate the travel time under assumed conditions. It is hoped to help improve the planning and designing of proper facilities with mixed traffic flow.
... This paper proposes a capacity model which assumes borrowed priority for bicycle stream at unsignalized intersections based on the addition鄄 conflict鄄flow ( ACF) procedure that was proposed by Brilon and Wu for application to unsignalized intersections [6 -8] . This procedure can easily han鄄 dle the intersection between different streams [9] . ...
Article
Full-text available
To investigate bicyclists' behavior at unsignalized intersections with mixed traffic flow, a bicycle capacity model of borrowed-priority merge was developed by the addition-conflict-flow procedure. Based on the actual traffic situation, the concept of borrowed priority, in which the major-road bicycles borrow the priority of major-road cars to enter the intersections when consecutive headway for major-steam cars is lower than the critical gap for minor-road cars, was addressed. Bicycle capacity at a typical unsignalized intersection is derived by the addition-conflict-flow procedure. The proposes model was validated by the empirical investigation. Numerical results show that bicycle capacity at an intersection is the function of major-road and minor-road car streams. Bicycle capacity increases with increasing major-road cars but decreases with increasing minor-road cars.
Article
This paper studies the effect dwelling mixed-traffic arterial Bus Rapid Transit (BRT-“Lite”) buses have on general traffic conditions and intersection capacity. Queue length and flow rate data collected from the busiest intersection along the bus route were used as proxies for quantifying traffic impacts. Two traffic scenarios were examined: (1) high volume oversaturated traffic; and (2) normal intersection conditions that were at or below saturation levels. Through use of linear regression models and a k-means clustering analysis—comparing traffic conditions before and after bus arrival as a function of green dwell time—it was found that BRT-Lite buses have no statistically significant impact on intersection performance or traffic capacity. This conclusion was further complemented by paired t-test results in which bus dwelling was found to result in an approximate change in both the average queue length and flow rate of only 1%. As a pilot study for the new BRT-Lite system, this research provides additional insights into the practical applicability of mixed traffic BRT systems and enriches the body of literature on related subjects.
Article
This study proposes a quantitative approach to evaluate the effects of mixed traffic flow on bus running times (except bus dwell times) near bus-stop areas based on linear regression and survival analysis theory. Research data were collected by video cameras at four bus stops in Nanjing, China. The application of the proposed methods with field data indicated that several factors would delay bus running times, i.e., car and nonmotor volume, bus dwell time, bus berth without violation, and nonmotor violations. In addition, the effect of bus lane-changing behavior on bus running times varied from section to section. Moreover, linear and parametric survival models were also developed to estimate bus running times, and both models can capture the effect of factors. The parametric survival models have better fitness performance than the linear models. However, the predictive performance of the two models are distinct at different sections and bus stops. The findings of influential factors and the proposed models could be considered in advanced bus information systems.
Article
Using the electricity consumption theory, we in this paper propose an electric vehicle's battery life model. The numerical results illustrate that the proposed model can qualitatively describe the impacts of the micro-driving behavior on the number of times that the electric vehicle's battery can cycle under two typical traffic situations. In addition, the numerical results show that the micro-driving behavior will influence the electric vehicle's battery life, i. e., under the smooth driving behavior, the number of times that the electric vehicle's battery cycles are stable and the number of the battery cycles will be increased, so the electric vehicle's battery life can be prolonged.
Article
After a detailed analysis of the characteristics of mixed traffic flow, the methods of Cellular Automata theory and rule description are used to build the pedestrian model, motor vehicle model and non-motor vehicle model. And then, the avoidance strategy is provided for motor vehicle as well as non-motor vehicle to avoid pedestrian. Finally, the simulation experiment prototype system of the mixed traffic flow is achieved by the means of system integration. The simulation results show that the system can better describe the complexity and uncertainty of the mixed traffic flow which provides decision-making basis for improving traffic flow at the intersections.
Conference Paper
Based on the analysis method of bus arrival time distribution from the departure station to the bus stop under the impact of an intersection signal, the expected capacity of road traffic near two different types of bus stops, including the curbside stop and the harbor bus stop, is obtained separately by using the proposed method. Also, a variety of factors that impact the actual road traffic capacity near bus stops are analyzed. The results show that the actual capacity of the road segment near the bus stop fluctuates with time, regularly in a cycle style, and the main factors include the departure frequency of the bus, variance of road travel time, the bus's dwelling time at the bus stop and the type of bus stop, all of which affect actual road traffic capacity near the bus stop. The road capacity near the bus stop can be released effectively by some measures, such as adjusting the departure interval times between buses at the terminal, and reforming the type of bus stop, etc.
Article
In this paper, we propose an extended car-following model and an electricity consumption model to study the effects of the road’s slope on the electric vehicle’s electricity consumption. The numerical results show that each electric vehicle’s electricity consumption increases with the uphill’s tilt angle and decreases with the downhill’s tilt angle. In addition, each electric vehicle’s electricity consumption increases with the uphill’s (downhill’s) length under a certain tilt angle.
Article
Full-text available
In many Chinese cities, traffic streams near bus stops differ from those of bus stops in developed countries. There are usually two lanes at a Chinese curbside stop: a nonmotorized lane and a motorized lane. Bus stops are often located on the nonmotorized lane. When a bus dwells at a curbside stop, nonmotorized vehicles, mainly bicycles, will move to the motorized lane. Thus, the presence of a stopped bus creates a temporary conflict between bicycles and cars, reducing road capacity. A road capacity model based on gap acceptance theory and queuing theory is presented for mixed traffic flow at the curbside stop. Traffic conditions are classified into two types: no stopped bus and presence of stopped bus at the curbside stop. The probabilities of no bus and presence of bus at a stop can be obtained by using the queuing model of bus streams. Under the former condition, the bicycle stream and the car stream have no conflict, and car capacity is not affected by the bicycle stream. Under the latter condition, the conflict between bicycles and cars on the motorized lane leads to an effect on car capacity by the bicycle stream. The effect on car capacity can be derived through gap acceptance theory. Car capacity at the curbside stop is a function of three types of traffic stream—buses, cars, and bicycles—and it may be applicable in traffic analysis and the design of bus stops in other developing Asian cities.
Article
Full-text available
A new simplified theoretical approach for the determination of capacities at unsignalized intersections has been developed on the basis of the method of additive conflict streams. This method is much easier to handle than the method of gap acceptance. It avoids many of the theoretical complications inherent in the method of gap acceptance, which, under certain circumstances, seems to be unrealistic. The new method has been developed for potential intersection configurations when one street has priority over the other. A calibration of the model parameters is given for German conditions. The new procedure can deal with shared lanes, short lanes, flared entries, and cases of so-called limited priority. For the estimation of traffic performance measures such as average delay and queue lengths, the classical methods can be applied.
Article
A new theoretical approach for the determination of capacity at all-way stop-controlled (AWSC) intersections is presented. This approach is based on the addition-conflict-flow method developed from graph theory. The new approach takes all the traffic streams into account, in contrast to existing procedures, which handle only the approaches. It allows a systematic and realistic analysis of the traffic process at AWSC intersections. The new approach can deal with most common lane configurations in the real world. A simple and more practical procedure is recommended. In practice, this simplified procedure can be used for single-lane approaches as well as approaches with separate left-turn traffic lanes. This procedure is verified and calibrated with measured data. For the calculation of capacity at AWSC intersections, general parameters, which are found by calibration, are proposed. The new procedure produces results more precise than the existing ones compared with the measured data. As a result, the maximum total capacities of AWSC intersections with single-lane approaches are found to be between 1,500 and 1,900 passenger cars per hour. The total maximum capacities of AWSC intersections with single-lane approaches and separate left-turn lanes are between 1,800 and 1,950 passenger cars per hour. Furthermore, the procedure shows that additional (left-turn) lanes significantly affect capacity increases only for asymmetric streetflow splits. The present procedure can easily be extended to AWSC intersections with multilane approaches.
Article
An iterative model for computing capacities at all-way stop-controlled (AWSC) intersections has been included in the new Highway Capacity Manual (HCM) 2000. The model is based on five saturation headway values, each reflecting a different level of conflict faced by the subject-approach driver. From this model the capacity and service time at any approach can be computed using iterative calculations. The model in the HCM is a so-called approach-based model, which only takes into account the conflicting cases among the approaches. The effect of turning streams or movements is not modeled in sufficient detail. In contrast, a theoretical, stream-based model for determination of capacities at AWSC intersections has been developed. This model is based on the addition-conflict-flow (ACF) method developed from graph theory and takes into account all the traffic streams at the intersection. This allows a systematic and realistic analysis of the traffic process at AWSC intersections. The computational procedure included in the model can be conducted without iterative calculation steps. The ACF and the HCM models for intersections with single-lane approaches were comprehensively validated, as well as a modified version of the HCM model that significantly enhances its features. The results of the validation indicate that the total capacity of an AWSC intersection with single-lane approaches based on the HCM model ranges between 1,450 and 1,550 passenger cars per hour (pc/h), whereas the total capacity based on the ACF model ranges between 1,600 and 2,000 pc/h. The modified HCM model yields total capacities ranging between 1,700 and 2,000 pc/h. The ACF model and the modified HCM model yield similar capacity results under normal traffic flow conditions.
Article
Bus use of urban roadways and past bus-capacity experience are reviewed. Field studies and bus simulation analyses were used to validate, update, and extend existing bus stop and berth capacity procedures. A 60% coefficient of dwell time variation was used to obtain new estimates of likely failure, and the maximum achievable capacity was based on a 25% failure rate. Capacity adjustment factors for skip-stop operation and right turns are derived. Service planning implications are identified. Bus lane capacities depend on how frequently the stops are placed, how long the buses dwell at each stop, traffic conditions and control systems along the bus lane or route, and whether buses can pass and overtake each other. Keeping dwell times and dwell-time variations to a minimum, providing multiple berth stops, and establishing skip-stop patterns will increase the bus and passenger capacity of bus lanes.
Article
When choosing the location and design for a particoular bus stop, sevcral alternatives are available. These alternatives include near-side, far-side, or midblock locations, and curbside or bus bay designs. Several studies have focused on choosing the optimum location of a bus stop for given situations: however, few have investigated the effects of bus stop design. The objective is to use computer simulation to determine how bus stop design influences traffic operation around a bus stop. Bus stop design analyzed included curbside, bus bay, open bus bay, and queue jumper. The results can be used to aidin the selection of a preferred bus stop design for a given location and traffic volume. The analysis was divided into two separate studies: cuebside versus bus bay/open bus bay. and queue jumper versus no queue jumper. The analysis consisted of investigating the relationships between variables such as travel time, speed, and traffic volume for given bus stop desigin and locations. The bus stop locations investigated in the curbsuide-bus bay/open bus bay study included midblock and far-side. Results indicated that the bus bay design provided the greatest benefit at traffic volumes of approximately 350 vehicles per hour per lane (vphpl) and above; however, notable advantages in vehicle speeds were also observed at 250 vphpl. Results from the queue jumper study revealed that the queue jumper design provided sinificant benefits at volumes above approximately 150 vphpl.
Article
This paper studies traffic features in a mixed traffic flow composed of motorized vehicles and nonmotorized vehicles near a bus stop within a completely different framework. Our work has taken the nonlane based behaviors of nonmotorized vehicles into account, which has seldom been considered in previous works. We have investigated flow rates as well as passenger transport capacity of the system. The traffic state diagram and the spatiotemporal pattern are presented as the inflow rates of motorized vehicles and nonmotorized vehicles change. A nonmonotonic change of passenger transport capacity is also identified at specific parameters. The influence of parameters such as stopped time of buses is discussed. These results are expected to shed some light on the management and design of urban mixed traffic systems.