Professional Documents
Culture Documents
Contents
Introduction
Reviewsomeoptimizationtechniques
Reviewsomepublications
SimpleapplicationusingExcel
Conclusions
INTRODUCTION
WhyOptimization?
AllahsaysintheQuran,"Hegiveswisdomto
whomHewills,andwhoeverisgivenwisdomis
certainlygivenalotofgood.Onlythepeopleof
understandingobservetheadvice.
(AlBaqara:269).
Introduction
Optimization
Objectivefunctions
Costminimization
BenefitMaximization
Technicalobjectives
Increasepower.Minimizeloss,maximizereleaseetc
Constraints
Physical
Economical
Sociological
Technical
ConventionalOptimizationModel
Linearprogramming
Nonlinearprogramming
Dynamicprogramming
SoftComputingTechniques
GeneticAlgorithms
NeuralNetworks
Fuzzylogic
REVIEWSOMEOPTIMIZATIONTECHNIQUES
Linearprogramming(LP)
Mostpopularoptimizationtechnique
Readilyavailablesolutionmethodology
Easyavailabilitiesofsoftwarepackages
Suitableforlargescalewaterresources
systems
Appliedwhentheobjectivefunctionand
constraintsarelinear
The general form of an LP model:
MAX (or MIN): c1X1 + c2X2 + + cnXn
Subject to:
LeonidKantorovich(1939)
LP Graphicalsolutionapproach
X2
261
250
200
180
150
boundary line of tubing constraint
12X1 + 16X2 = 2880
100
Feasible Region
50
0
0
50
100
150
174
200
240
250
X1
LP Enumeratingthecornerpoints
X2
250
o.f.v. = $54,000
(0, 180)
200
o.f.v. = $64,000
(80, 120)
150
o.f.v. = $66,100
$15,000
(122, 78)
100
o.f.v. = $0
o.f.v. = $60,900
50
(174, 0)
(0, 0)
0
0
50
100
150
200
250
X1
Nonlinearprogramming
AnNLPproblemhasanonlinear
objectivefunctionand/orone
ormorenonlinearconstraints.
NLPproblemsareformulated
andimplementedinvirtually
thesamewayaslinear
problems.
Themathematics(calculus)
involvedinsolvingNLPsisquite
differentthanforLPs.
AlbertW.Tucker
HaroldW.Kuhn
1950
PossibleOptimalSolutions
objectivefunction
levelcurve
objectivefunction
levelcurve
optimalsolution
optimalsolution
Feasible
Region
Feasible
Region
nonlinearobjective,
linearconstraints
linearobjective,
nonlinearconstraints
objectivefunction
levelcurve
objectivefunction
levelcurves
optimalsolution
Feasible
Region
nonlinearobjective,
nonlinearconstraints
optimalsolution
Feasible
Region
nonlinearobjective,
linearconstraints
Dynamicprogramming
RichardE.Bellman(1940)
GeneticAlgorithms
GAprosedure
AnalyzetheproblemanddetermineO.F.,decision
variables,Searchspaceandconstrains
Generateinitialpopulation(possiblesolutions)
Encodethepopulation(Binary,Value,Permutation
andTree)
Evaluatethefitnessofeachindividualinpopulation
Repeatuntilterminationconditionsatisfied:
Selection:Selecttheindividualswithgreaterfitnessfor
reproduction
NilsAall Barricelli (1954)
Crossover:Breednewindividualsthroughcrossover
Mutation:Applyprobabilisticmutationonnewindividuals
Formanewpopulationwiththeseoffspring.
Terminate:whengoalconditionachieved
GeneticAlgorithms(cont.)
Generatedpopulation
1
Encoding
Evaluatingthefitness andselection
25%
58%
7%
93%
65%
1%
25%
58%
93%
65%
Crossover
Mutation
Repeatuntilsuccessconditionachieved
Encoding,Evaluating,Selecting,Crossover,Mutation
REVIEWSOMEPUBLICATIONS
10reservoirsindifferentriverbasins
watersupply,floodprotection,
hydropowergeneration
33demandcenters
Objectivefunction
Fillingrules(M)thatminimize
hydropowershortagein(R)?
LinkMIKEBASINandNSGAII
HydrologicalData
(Inflow,irrigationdemands,
evaporation,waterlevelsat
bothreservoirsandRoseires
ratingcurves.)
SIMPLEAPPLICATIONUSINGEXCEL
)(
)./.(
)(
)(
) (
)(
CONCLUSIONS