selfadaptive differential evolution applied to realvalued antenna and microwave design problems外语英文电子书

《selfadaptive differential evolution applied to realvalued antenna and microwave design problems外语英文电子书》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、1286IEEETRANSACTIONSONANTENNASANDPROPAGATION,VOL.59,NO.4,APRIL2011Self-AdaptiveDifferentialEvolutionAppliedtoReal-ValuedAntennaandMicrowaveDesignProblemsSotiriosK.Goudos,Member,IEEE,KatherineSiakavara,Member,IEEE,TheodorosSamaras,Member,IEEE,EliasE.Vafiadis,Member,IEEE,andJohnN.Sahalos,Li

2、feFellow,IEEEAbstract—Particleswarmoptimization(PSO)isanevolutionarybeenappliedtomicrowavestructures[26]–[28],antennadesignalgorithmbasedonthebirdfly.Differentialevolution(DE)isa[29]–[35],signaloptimization[36]andmicrowaveimagingap-vectorpopulationbasedstochasticoptimizationmethod.Thefact

3、plications[37]–[44].thatbothalgorithmscanhandleefficientlyarbitraryoptimizationDEproducedbetterresultsthanPSOonnumericalbench-problemshasmadethempopularforsolvingproblemsinelectro-magnetics.Inthispaper,weapplyadesigntechniquebasedonamarkproblemswithlowandmediumdimensionality(30andself-ada

4、ptiveDE(SADE)algorithmtoreal-valuedantennaand100dimensions)[45].However,onnoisytestproblems,DEwasmicrowavedesignproblems.Theseincludelinear-arraysynthesis,outperformedbyPSO.In[46]acomparativestudybetweenDEpatch-antennadesignandmicrostripfilterdesign.ThenumberandPSOvariantsispresentedforth

5、edesignofradarabsorbingofunknownsforthedesignproblemsvariesfrom6to60.Wecomparetheself-adaptiveDEstrategywithpopularPSOandmaterials(RAM).Thenumberofproblemdimensionswas10DEvariants.Weevaluatethealgorithms’performanceregardingandDEoutperformedthePSOvariantsintermsofconvergencestatisticalre

6、sultsandconvergencespeed.Theresultsobtainedforspeedandbestvaluesfound.Theshapereconstructionofaper-differentproblemsshowthattheDEalgorithmsoutperformthefectlyconducting2-DscattererusingDEandPSOispresentedPSOvariantsintermsoffindingbestoptima.Thus,ourresultsin[40],[44].Alsobothalgorithmsha

7、vebeenappliedto1-DshowtheadvantagesoftheSADEstrategyandtheDEingeneral.However,theseresultsareconsideredtobeindicativeanddonotsmall-scaleinversescatteringproblems[43].Inthesecases,DEgenerallyapplytoalloptimizationproblemsinelectromagnetics.outperformedPSO.In[47]acomparison