Upload
ayenigba-sogo-emmanuel
View
219
Download
0
Embed Size (px)
DESCRIPTION
Academic
Citation preview
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 1/13
EVOLUTIONARY ALGORITHMS (EA) AS OPTIMIZATION
TOOLS
(CLASS SEMINAR)CVE 697
BY
AYENIGBA, SOGO EMMANUEL
08/30GB096
AN
IGE, E!AR OLU!ASEUN0"/30GB0#7
EPARTMENT O$ CIVIL ENGINEERING
UNIVERSITY O$ ILORIN, ILORIN
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 2/13
ABSTRACT
• The need for nding optimal solutions in problemscomes with the extremities of designing formaximum cost or minimum reliability. Henceoptimi!ation methods are of great importance."#er the years se#eral research and applications
in optimi!ation has been seen. This study focuseson one of the non$traditional optimi!ationmethods which is %#olutionary Algorithms &%As'.%As o(er practical ad#antages to researchers
facing di)cult optimi!ation problems due to itsrobust response to changing circumstances*exibility among other facets. The study alsohighlights the principle behind %As which is the+arwin,s principle of %#olution. The general
characteristics which include selection and
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 3/13
/TR"+0CT"/
• "ptimi!ation – +enition
– 1athematical and %ngineering"ptimi!ation
– History
– /on$traditional method of solution
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 4/13
+%2/T"/ "2 %As
• %#olutionary algorithms &%As' arepopulation$based metaheuristicoptimi!ation algorithms that use
biology$inspired mechanisms li3emutation crosso#er naturalselection and sur#i#al of the ttest
in order to rene a set of solutioncandidates iterati#ely
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 5/13
BASC 4R/C45% 2R"1/AT0R%
• +arwin,s principle of naturalselection
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 6/13
-%/%RA5 CHARACT%RSTCS "2 %As
• Set of solution candidates
• Selection process
•
Recombination and mutation
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 7/13
-%/%RA5 SCH%1% "2 %A
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 8/13
-%/%RA5 SCH%1% "2 %Acontd6
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 9/13
T74%S "2 %As
• -enetic algorithms
• %#olutionary strategy
•
%#olutionary programming• -enetic programming
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 10/13
C"/C50S"/• "ptimi!ation methods are of great importance. The
need for nding optimal solutions in problems comes
with the extremities of designing for maximum cost orminimum reliability.
• "#er the years se#eral research and applications inoptimi!ation has been seen. This study showed that
there were se#eral methods that had been used intrying to sol#e optimi!ation problems includingtraditional &mathematical' and non$traditional methods.
• The study focused on non$traditional optimi!ationmethods which is %#olutionary Algorithms &%As'. %As
o(er practical ad#antages to researchers facing di)cultoptimi!ation problems due to its robust response tochanging circumstances *exibility among other facets.
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 11/13
C"/C50S"/ contd6.
• %As was dened as a population$based
metaheuristic optimi!ation algorithms that usebiology$inspired mechanisms li3e mutationcrosso#er natural selection and sur#i#al of thettest in order to rene a set of solution
candidates iterati#ely The study highlighted theprinciple behind %As which is the +arwin,sprinciple of %#olution. The general characteristicswhich include selection and recombination were
also discussed. Among the se#eral types of %As-enetic algorithm %#olutionary strategy -eneticprogramming and %#olutionary programmingwere discussed.
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 12/13
R%C"11%/+AT"/
• 2urther study to be carried out withthe aim of de#eloping %#olutionaryAlgorithms procedure.
• %As should be gi#en more classroomhours so as to enable the studentsbenet more from its rich resource.
7/21/2019 Evolutionary Algorithms (Ea) as Optimization Tools
http://slidepdf.com/reader/full/evolutionary-algorithms-ea-as-optimization-tools 13/13
THA/8 7"02"R 5ST%//-