Usingtargetedrepresentation.
Thesetoftargetedbuildingblocksevolvedduringthisprocessisthenusedasaninitialsetofbuildingblocksduringthesecondexperimentwhenweproducethedesignswithmaximaltotalareaofholesinsideandmaximalnumberofconnec-tionsbetweentheseholesinsidethestructure.Herethe tnessesareclosetobut
Abstract. This paper presents a formal approach to the evolution of a representation for use in a design process. The approach adopted is based on concepts associated with genetic engineering. An initial set of genes representing elementary building blocks
14JohnS.GeroANDVladimirA.
Kazakov
Figure11.(a)Anexampleoftheevolutionarypathsintheevolutionofacomplexbuildingblock,(b)someofthedesignsproducedusingthesetofevolvedcomplexblocks.
Abstract. This paper presents a formal approach to the evolution of a representation for use in a design process. The approach adopted is based on concepts associated with genetic engineering. An initial set of genes representing elementary building blocks
EvolvingBuildingBlocksforDesignUsingGeneticEngineering
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
015TOTAL FRACTION OF COMPLEX GENES020406080100GENERATIONS120140160180
Figure12.Thefractionofcompositebuildingblockinthetotalpoolofbuildingblockusedtoas-semblethepopulationvs.generationnumber.Inthisexperimenttheobjectivefunctionisthenum-berofclosedholesandthenumberofconnectionbetweentheclosedholesinsidethestructure.Theinitialsetofbuildingblocksisinheritedfromthe rstexperimentandisthetargetedrepresentation.notthesameasthoseusedtoevolvethetargetedrepresentation.Thisexperimentisusedtotestwhetherthetargetedrepresentationislikelytobeusedmorethantheoriginal,elementarybuildingblocks.Ifthetargetedrepresentationisusedratherthantheelementarybuildingblocksthenwehaveachievedourgoalofevolvingarepresentationcanbeusedtoproducedesignswhichexhibitdesiredcharacterist-icsmorereadily.TheresultsareshowninFigures12and13.Onecanseethatthefractionofthecompositebuildingblocksusedtoproducethesedesignsreachesthesaturationlevelduringthe rstfewiterations.Thevisibleredistributionsofthepopulationbetweenthecompositebuildingblocksof5,6and7-complexitiesarepurelysuper cial-thisredistributionoccursbetweenthesamecompositebuildingblockswhicharepresentinalltheserepresentations.Evolutionoftherepresenta-tiondoesnotoccurduringthisexperiment-nonewcomplexbuildingblockwereevolved.Thiscanbeinterpretedasanindicationofclosenessofthetargetedrep-resentationsforbothproblems.Soifthetargetedrepresentationisevolvedforonesetofobjectivesthenitcanbeusefullyappliedtoanyoftheobjectivesetswhichareonlyslightlydifferenttoit.
Effectsofincompleteevolution
Inthisexperimentwerepeatthe rstiterationbutstoptheevolutionprematurelyafteronly60generations.Afterthiswerepeattheseconditerationusingtheevolvedincompletesetofcompositebuildingblocks.TheresultsareshowninFigures14
Abstract. This paper presents a formal approach to the evolution of a representation for use in a design process. The approach adopted is based on concepts associated with genetic engineering. An initial set of genes representing elementary building blocks
16
0.9JohnS.GeroANDVladimirA.Kazakov0.8
0.7
FRACTION OF COMPLEX GENES 0.60.51-COMPLEXITY2-COMPLEXITY3-COMPLEXITY4-COMPLEXITY5-COMPLEXITY6-COMPLEXITY7-COMPLEXITY8-COMPLEXITY0.40.3
0.2
0.1
0020406080100GENERATIONS120140160180
Figure13.Thefractionofcompositebuildingblockswithdifferentcomplexitiesinthetotalpoolofthebuildingblockusedtoassemblethepopulationvs.generationnumberintheexperiment.
1
0.9
TOTAL FRACTION OF COMPLEX GENES 0.80.7 0.60.50.40.3
0.2
0.10102030GENERATIONS405060
Figure14.Thefractionofcompositebuildingblockinthetotalpoolofbuildingblockusedtoas-semblethepopulationvs.generationnumber.Inthisexperimenttheobjectivefunctionisthenumberofclosedholesandthenumberofconnectionsbetweentheclosedholesinsidethestructure.Theini-tialsetofbuildingblocksisinheritedfrom rstiterationwhichhasbeenprematurelyterminatedatgeneration60.
and15.Inthiscasetheevolutionoftherepresentationcontinuesforaboutafur-ther10generationsandweendupwiththesamesetofevolvedcompositebuildingblocks.Thesaturationofthepopulationwiththecompositebuildingblocksisalsocompletedafterthese10generations.Thus,onecanstarttoevolvearepresenta-
Abstract. This paper presents a formal approach to the evolution of a representation for use in a design process. The approach adopted is based on concepts associated with genetic engineering. An initial set of genes representing elementary building blocks
EvolvingBuildingBlocksforDesignUsingGeneticEngineering
0.9170.8
0.7
FRACTION OF COMPLEX GENES 0.60.51-COMPLEXITY2-COMPLEXITY3-COMPLEXITY4-COMPLEXITY5-COMPLEXITY6-COMPLEXITY7-COMPLEXITY8-COMPLEXITY0.40.3
0.2
0.1
00102030GENERATIONS405060
Figure15.Thefractionofcompositebuildingblockswithdifferentcomplexitiesinthetotalpoolofthebuildingblockusedtoassemblethepopulationvs.generationnumberinthethirdexperiment.tionforonesetofobjectivesandthencontinueitforanothercloselyrelatedsetofobjectives.