197
1 INTRODUCTION
.
A marine aid to navigation is a device or system
external to vessels that is designed and operated to
enhance the safe and efficient navigation of vessels
and/orvesseltraffic(IALA,2001).
The AtoNs system marks a navigational channel,
so the risk level of the AtoNs not only reflects the
navigationservicelevelof thi
s system also indicates
theshipnavigationsafetylevelinthischannel.After
the real‐time risk assessment of the AtoNs system,
somesuggestionscanbe given tothe administrators
toimprovethenavigationservicelevelandguarantee
theshipnavigationsafety.
During the whole lifecycle of an aid, the risk
assessmentforma
rineaidsisalwaysnecessary. But,
atpresent,theAtoNsriskassessmentmainlydepends
on the subjective judgment of the administrators or
experts, or relies on some indirect suggestions from
relatedaidstonavigationinformationsystems.These
methods can not produce real‐time, accurate and
systematicassessmentresults.So,ma
kingfulluseof
the existing advanced information systems in the
administration department, extracting real‐time,
accurate and comprehensive data from the systems,
thencarryingonreal‐timeriskassessmentforaidsto
navigation in chosen channel, is very significant to
improvetheserviceleveloftheAtoNssystem.Bythi
s
means,theriskassessmentcanrunthroughthewhole
Real-time Risk Assessment for Aids to Navigation
Using Fuzzy-FSA on Three-Dimensional Simulation
System
J
.Chen,T.Chen,C.Shi&D.Jia
M
erchantMarineCollege,ShanghaiMaritimeUniversity,Shanghai,China
ABSTRACT:TheriskleveloftheAidstoNavigation(AtoNs)canreflecttheshipnavigationsafetylevelinthe
channeltosomeextent.Inordertoappreciatetheriskleveloftheaidstonavigation(AtoNs)inanavigation
channel and to provide some decision‐ma
king suggestions for the AtoNs Maintenance and Management
Department,theriskassessmentindexsystemoftheAtoNswasbuiltconsideringtheadvancedexperienceof
IALA. Under the Formal Safety Assessment frame, taking the advantages of the fuzzy comprehensive
evaluationmethod,thefuzzy‐FSAmodelofriskassessmentforaidstonavigationwasest
ablished.Themodel
wasimplementedfortheassessmentofaidstonavigationinShanghaiareabasedonthe aids tonavigation
three‐dimensionalsimulationsystem.Thereal‐timedatawereextractedfromtheexistinginformationsystem
of aids to navigation, and the real‐time risk assessment for aids to navigation of the chosen channel was
performed on pla
tform of the three‐dimensional simulation system, with the risk assessment software.
Specifically, the deep‐water channel of the Yangtze River estuary was taken as an example to illustrate the
generalassessmentprocedure.Themethodproposedpresentspracticalsignificanceandapplicat
ionprospect
onthemaintenanceandmanagementoftheaidstonavigation.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 8
Number 2
June 2014
DOI:10.12716/1001.08.02.04
198
procedure of an aid including plan, placement,
construction, operation, maintenance and
management,andcandecreasesubjectivejudgments,
which can provide more useful, comprehensive and
real‐time information and advices for the
administrators.
The index system forAtoNs risk assessment was
established by exploring experience from IALA and
consulting many related experts. Then, under the
framework of the Formal Safety Assessment (FSA)
method,thefuzzycomprehensiveassessmentmethod
was int
roduced into the work to build a Fuzzy‐FSA
model to realize the risk assessment for the AtoNs.
Besides, the real‐time risk assessment for aids to
navigation of the chosen channel was performed on
pla
tformof the three‐dimensional simulation system
of aids to navigation, with the risk assessment
software. Specifically, the deep‐water channel of the
Yangtze River estuary was taken as an example to
illustratethegeneralassessmentprocedure.
2 INDEXSYSTEMFORATONSRISK
ASSESSMENT
Amarine aid tonavigation,asan individual pa
rtin
the channel, is very easily influenced by external
environment, including the navigation vessel
conditions, traffic conditions, channel conditions,
hydrological conditions, meteorological conditions
andsoon.Frequentabnormalconditionshappenedto
amarineaidmaybe:damage,abnormallight,shifting,
lost,etc.Theseallwillhaveobviousharmful effect
sto
the navigation service level for the ships and the
navigation safety, even may lead to ship collision,
shipgroundingorcollisionbetweenshipsandbuoys.
Therefore, the risk assessment for marine aids is
necessary and covers so many factors, such as
navigation ships, traffic, channel, environment,
accidents history and so on. After drawing lessons
from the <IALA Aids to Navigation Gui
de> and
consultingsomerelatedexpertsonaidstonavigation
maintenance and management (from Aids to
Navigation Administration Departments, Aids to
Navigation Plan and Design Departments, Pilot
StationsandAidstoNavigationRepairStations),also,
considering the operability of the real‐ti
me
assessment on the aids to navigation three‐
dimensionalsimulationsystem,theindexsystemfor
AtoNsriskassessmentwasestablishedinTable1.
Besides,somefactorshavingrelationshipswiththe
riskofAtoNsmaybeexcludedfromtheindexsystem
becauseoftheirtinyandnegligibleeffectscomparing
to other fact
ors in the system, for example “channel
length”in“waterwayconfiguration”,orthemeanings
ofthemhavebeenembodiedintootherfactorsinthe
system, for example “ship size” embodied into
“trafficmix”viastatistics.Inaword,thebuiltindex
system for AtoNs risk assessment is relatively
completeandfeasible.
Table1.IndexsystemforAtoNsriskassessment
_______________________________________________
RiskassessmentforAtoNs
_______________________________________________
1. Traffic Deepdraught
Volume Shallowdraught
Commercialfishingvesselsandother
boats
Hazardcargoes
_______________________________________________
2. Ship Trafficmix
Traffic Trafficdensity
Conditions Shipspeed
_______________________________________________
3. Navigational Visibility
conditions Wind
Currentandwave
Obstructionscondition
Aidstonavigationcondition
_______________________________________________
4. Waterway Channelwidth
configuration Channelcurvature
Waterwaycomplexity
Channeldepth
Channelstructure
_______________________________________________
5. Accident Accidentfrequency
conditions Injuriestopeople
Propertydamage
Hazardousmaterialrelease
Emergentrescueequipmentcondition
Emergentrescuesystemlevel
_______________________________________________
3 FUZZY‐FSAMODEL
3.1 IntroductionofFormalSafetyAssessment(FSA)
FSA is a structured and standardized safety
assessment method. In the 20th century, for
promoting and improving the maritime safety, the
International Maritime Organization (IMO)
encourages Member States to apply this advanced
safetyassessmenttospecialresearchonthesafetyof
vessels. So far, FSA has been widely applied int
o
safetyrulemaking,shipdesignandshipmanagement
and other related fields. It provides some decision‐
making proposals to improve the navigation safety
level and reduce or avoid marine risk. FSA method
has five formal steps shown in Fig 1, including
i
dentity risks/hazards, assess risks, specify risk
controloptions,makeadecisionandtakeaction.
Comparingtoothermethods,thestepsinFSAare
much more reasonable and comprehensive. Also, it
can be integrated into some comprehensive
evaluationmethodstoanalyzetheriskandinfluence.
This method will propose corresponding decision‐
ma
king suggestions from both the quantitative and
qualitativeangle.Then,theevaluationresultswillbe
muchmorescientific,makingthesuggestionsonrisk
controlmorepracticalandfeasible.
1. Identity
risks/hazards
5. Take action2. Assess risks
3. Specify risk
control options
4. Cost-benefit
analysis
Figure1.StepsofFSAmethod
199
3.2 IntroductionofFuzzy‐FSAmodel
Fuzzy comprehensive evaluation method (Fuzzy)
appliesfuzzytransform principle andthe maximum
membership degree law, and considers factors
associated with the target to make a comprehensive
evaluation. The evaluation results can reflect the
actual conditions of the evaluated target
comprehensively after analyzing from multi‐fact
or
andmulti‐level.Now,Fuzzymethodhasbeenwidely
usedinvariousfields.
The general steps of the Fuzzy comprehensive
evaluationmethodareasfollows:
Building the risk evaluation index system with
hierarchical structure based on the characteristics
ofthetargettobeevaluated;
Determiningoftheweightsetoftheindexsystem
byexpertconsultingandAHPmethod;
Building of the evaluation ma
trix of each factor
accordingtothedeterminedquantitativestandard;
Calculatingthefinalfuzzyrelationmatrixforthe
targetbasedontheweightset;
Obtaining the quantitative and qualitative
evaluation results through the defuzzificat
ion of
theresults.
Fuzzy‐FSAmodelisbuiltbyintroducingtheFuzzy
comprehensive evaluation method under the
framework of FSA method. The qualitative and
quantitative risk level will be achieved by building
the index system and fuzzy assessment. Fuzzy
evaluationmethodisobviousinsteps1,2,3andstep
5 of FSA method. Because of considering too ma
ny
economicandpoliticalfactors,thestep4ofFSA,cost‐
benefit analysis is excluded here. FSA provides
overall assessment ideas while the Fuzzy method
providesevaluation techniqueand evaluation index.
These two methods combined together to make the
evaluationideamoreclearandtheevaluationresults
morepersua
siveandfeasible.Fig2showsthegeneral
stepsofFuzzy‐FSAmethod.
Figure2.GeneralstepsofFuzzy‐FSAmethod
3.3 QuantitativeStandardofEachFactor
Inordertorealizethereal‐timeriskassessmentofthe
AtoNs, each factor in the index system should be
analyzed to determine one corresponding
quantitative standard. The standard must be so
accurate as much as possible that can represent the
actual meaning of the fact
or, and also should be
feasiblemakingsurethevalueofthefactoriseasyto
get.
The meanings of some factors in Table 1 are
obviousandclear,sothequantitativestandardsof
themcan be determined easily according to their
literal meanings, for example, “traffic mix”,
“current and wave”, “channel depth” and
“accidentfrequency”.
While some fact
ors can be quantified into a
percentage,forexample,using“theproportionof
large ships per day (%)” to indicate the “Deep
Draught”,using“theproportionofover‐speeding
ships per day (%)” to indicate the “Ship Speed”,
the same to the other three fact
ors: “Shallow
draught”, “Commercial fishing vessels and other
boats”,“Hazardcargoes”.
While some factors are qualitative, such as,
“channel structure”, “aids to navigation
condition”, “hazardous material release”,
“emergent rescue equipment condition” and
“emergentrescuesystemlevel”.Thesefactorscan
be quantified by means of a score (from 0~10)
determinedbytheadministratorsorinvest
igation
results.
Then,“Injuriestopeople”and“propertydamage”
canbequantifiedbystraightdata,whichare“the
number of people injured” and “the economic
lost”respectively.
Because of lacking of intuitive and appropriate
judgment criteria, or no direct data, the other
fact
ors (visibility, wind, channel width, channel
curvature, obstructions condition, waterway
complexity)cannotbequantifiedaccordingtothe
superficial meanings, needs further analysis.
Taking the “visibility” as an example, even one
channelmay havedifferentvisibilitiesevery day,
so it is very inconvenient to collect the data and
unsuitabletocomparebetweendifferentchannels
ifjustusingthe“t
hedistanceof the visibility”to
qualify it. Therefore, after further analyzing the
factor and making reference to related research,
“the days per year with the visibility less than
1km” is determined as the quantitative standard
for“visibility”.Also,thequantitativestandardsof
“wind”,“channeldepth”,“channelcurvature”and
“wat
erwaycomplexity”allhavebeendetermined
bythismeans.
Thequantitativestandardofeveryfactorisshown
inTable2.
200
Table2.Quantitativestandardofeachfactorandtheweightindex
__________________________________________________________________________________________________
RiskassessmentforAtoNsQuantitativeStandardWeight
__________________________________________________________________________________________________
1. Traffic Deepdraught Theproportionoflargeshipsperday(%) 0.055
Volume ShallowdraughtTheproportionofsmallshipsperday(%) 0.037
Commercialfishing TheproportionofCommercialfishingvesselsandotherboatsperday(%)0.037
vesselsandotherboats
HazardcargoesTheproportionofshipswithhazardcarg
oesperday(%)0.055
__________________________________________________________________________________________________
2. Ship Trafficmix Averageshipnumbersperday 0.102
Traffic TrafficdensitySmoothnesslevel:Score(0~10) 0.077
Conditions ShipspeedTheproportionofover‐speedingshipsperday(%)0.077
__________________________________________________________________________________________________
3. Navigational Visibility Thedaysperyearwiththevisibilitylessthan1km 0.021
conditions Wind Thedaysperyearwithwindstrongerthan6grade 0.038
Currentandwave Speedofcrosscurrent(m/s) 0.034
Obstructionscondition Distancetothechannelcenterline(m) 0.021
Aidstonavigation Score(0~10) 0.034
condition
__________________________________________________________________________________________________
4. Waterway Channelwidth Thelengthofthelargestship/themostnarrowwidthofthechannel 0.097
configurationChannelcurvature Thelargestturningangleofthechannel(Deg) 0.049
Waterwaycomplexity Numbersoftraffic specialpointsinthechannel/channellength0.025
ChanneldepthChanneldepth(m) 0.021
Channelstru
cture Score(0~10)0.021
__________________________________________________________________________________________________
5. Accident Accidentfrequency AverageAccidentfrequencyperyear 0.034
conditions Injuriestopeople Thenumbersofpeopleinjured 0.033
Propertydamage Theeconomiclost 0.033
Hazardousmaterial Pollutiondegree:score(0~10) 0.033
release
Emergentrescue score(0~10) 0.033
equipmentcondition
Emergentrescuescore(0~10) 0.033
sy
stemlevel
__________________________________________________________________________________________________
3.4 WeightofEachFactor
Theweightofeachfactorisalsoveryimportanttothe
risk assessment results. For making sure of the
accuracy and acceptability of the assessment results,
onthebasisoftherelatedprojectfromShanghaiAids
toNavigationAdministrationDepartment,theweight
indexwasacquiredbyint
egratingDelphimethodand
Analytic Hierarchy Process (AHP) method. During
the project, the expert questionnaire was complied
and sent to 48 experts from Aids to Navigation
Administration Departments, Aids to Navigation
PlanandDesignDepartments,PilotStationsandAids
to Navigation Repair Stations and some related
departments. The questionnaires were all replied
ba
ck.Fullyconsideringalltheexperts’commentsand
suggestions, the judgment matrix based on AHP
method was established. The weight of each factor
was calculated by square root rule, and passed the
consistencycheck.Thedetailedcalculationprocedure
willnotberepeatedhere.
Besides,theweightindexshouldalsobedynamic
and adjust
able in coincidence with the changes in
channel or port or government policy. Because with
the development of port and channel and the
continuousconstruction of AtoNs, the risk may face
willalsobechanged.Theweightindexisalsoshown
inTable2.
Owningtolimit
edtimeandresources,thenumber
ofthesentquestionnairesmaybenotenough.Butthe
experts chosen are all very experienced and
representativeinthefieldinChina.Theirpersuasive
comments and suggestions can generally guarantee
theobjectivityandaccuracyoftheassessmentresults.
4 REAL‐TIMERISKASSESSMENTFORATONS
USINGFUZZY‐FSAONTHEPLATFOMOFAIDS
TONAVIGATIONTHREE‐DIMENSIONAL
SIMULATIONSYSTEM
Till now, the AtoNs administrators mastering and
assessing the AtoNs risk mainly resorts to
measurements by pra
ctical ship trail. It is time‐
consumingandcostexpensive.Themostimportantis
that the results by technical measurements only can
reflectthecondit
ionsoftheaidsobservedunderthe
environment condition at that time, but not
comprehensive and real‐time. Besides, the results
measuredareeasilyaffectedbythetechnicallevel,the
distance from the target, the testers and the
environmentconditionsandsomeotherfactors.
Now, China Aids to Navigation Administration
Departmentshavebeeneq
uippedmanyusefulAtoNs
information system to help mastering the real‐time
conditions of the AtoNs in the
waterareasunderthejurisdiction,suchasAISAtoNs
system, Aids to Navigation Three‐dimensional
Simulation System, Aids to Navigation Telemeter
andTelecontrol System, and so on. Especially, the
Aids to Navigation Three‐dimensional Simulation
System is a comprehensive system, int
egrating with
other information systems, which can display the
surrounding conditions in real time, including the
aids to navigation and nearby ships’ information
underthepre‐setenvironmentconditions.So,itisan
optimal platform to realize the real‐ti
me risk
assessmentforaidstonavigation.
201
Embed
Display
Feedback
Aids to Navigation
Three-dimensional
Simulation System
AIS information
Aids to Navigation Telemeter
and Telecontrol System
Others
Real-time
AtoNs data
Channel Data
External Database
AtoNs Risk
Assessment Module
Index
System
Fuzzy-
FSA
Model
Results
&
Suggestions
Figure3.GeneralProcedureofthereal‐timeriskassessmentforaidstonavigation
Table3.Dataclassificationanddatasource
__________________________________________________________________________________________________
Classification DataneededDatasource
__________________________________________________________________________________________________
Basicfactdata NavigationalconditionsStatisticsMaterial, suchassailingguidelines
ChannelConditionsChanneldatabaseandthesupplementinthesimulationsystem
__________________________________________________________________________________________________
Real‐timedata TrafficVolumeconditions AISdatainthree‐dimensionalsimulationsystem
ShiptrafficconditionsAISdatainthree‐dimensionalsimulationsystem
Accidentfrequency;Injuriestopeople; Accident statisticsmaterials
Propertydamage;
Hazardousmaterialrelease
__________________________________________________________________________________________________
Externaldata AidstonavigationconditionEvaluationresultsformtheadministrators
Emergentrescueequipmentconditions; Actualconditionsandexpertjudgments
Emergentrescuesystemlevel
__________________________________________________________________________________________________
4.1 GeneralProcedureofthereal‐timeriskassessmentfor
aidstonavigation
If all the values of the influencing factors of aids to
navigation risk were input or determined manually,
theriskassessmentsystemwouldbemeaningless.So,
anAtoNsriskassessmentmodulewasdevelopedon
the platform of the Aids to Navigation Three‐
dimensional Simulation System. It was embedded
int
o the simulation system, also, can run
independently.
The real‐time risk assessment for aids to
navigation is realized by extracting the needed real‐
time data from the aids to navigation three‐
dimensional simulation system. And also, some
additional external database and expert experience
knowledgewillbe supplemented to the system. The
data can be updated real‐ti
mely according to the
chosen targets and assessment conditions. In this
module, the important real‐time information was
extractedtodeterminethevalueofthefactorsinthe
built index system, then, activating the fuzzy‐FSA
modeltocarryouttheriskassessmentforchosenaid
or channel. The results can reflect the real ti
me risk
levelof the aidstonavigation inthe channel,which
can provide more helps for the aids to navigation
administrators and be useful for improving the
navigation service level and ship navigation safety
level.And,theresultscanbedisplayedonthethree‐
dimensionalsimulationsystem,alsocanbesavedas
anexcelta
bletobeprinted,whichisveryconvenient
tocompareandanalyzeinfuture.
Thegeneralprocedureoftheassessmentisshown
inFig.3.
4.2 Preparationofdata
The data needed in the assessment includes the
va
lues of the factors and their weights. Then,
according to the built risk assessment index system
and the quantitative standards, the value of each
factor can be classified into three types, which are
basicfactdata,real‐timedataandexternaldata(Table
3).Itisobvioustha
tthenavigationalconditionsand
the channel conditions of one channel are often
constant in a long time, while the ship conditions,
ship traffic conditions and the accidents conditions
aremainlydynamic.Besides,someotherdata, suchas
the aids to navigation service, emergent rescue
equipment condit
ions and emergent rescue system
level, rely on external input or judgment. This
classificationishelpfultothedatapreparationbefore
riskassessment.
Differenttypesofdataarefromdifferentsources,
shown in Table 3. The Aids to Navigation Three‐
dimensional simulation system, as the foundation
pla
tform, is certainly the first data source.
Particularly, its bottom aids to navigation database
and channel database will provide some precious
datafortheriskassessment.Meanwhile,thedatabase
can be supplemented and perfect according to
requirements.Besides,thebasicfactdataandthereal‐
time data can be complied int
o a database for the
directuseinthethree‐dimensionalsimulationsystem.
And,thisishelpfultodatamodificationandupdate,
guaranteeing the timeliness and accuracy of the
evaluationresult.
4.3 DesignResultsoftheReal‐timeriskassessment
module
BasedontheECDIS,usingVisualC++andDatabase
Management technology, the real‐ti
me risk
assessment system for aids to navigation on Three‐