简介:Tosolvetheproblemoflowefficiencyinpiperoutingdesign,animprovedgeneticalgorithmbasedapproachisproposed.Topresentthisapproach,thepapermainlydescribesagenerationmethodofnodesconsideringthesafetydistanceofpipesandthedirectionalconstraintsatterminals,thedefinitionofadoublecodingtechnique,thecollisiondetectionmethod,theconceptofenergyandthedefinitionoffitnessfunctions.Thesimilaritydetectionisintroducedtopreventclosebreedinginthecrossoveroperator,theselectionpressureiscontrolledaccordingtotheevolutionsituationandaheuristicmutationmethodisusedtoboosttheevolution.Simulationcaseshowsthatthisapproachismorepracticalandcansatisfydifferentdesignrequirementsbychangingalgorithmparameters.
简介:Inordertoovercomethepoorlocalsearchabilityofgeneticalgorithm,resultinginthebasicgeneticalgorithmistime-consuming,andlowsearchabilityinthelateevolutionary,weusethegraycodinginsteadofbinarycodingatthebeginningofthecoding;weusemulti-pointcrossovertoreplacetheoriginalsingle-pointcrossoveroperation.Finally,theexperimentshowsthattheimprovedgeneticalgorithmnotonlyhasastrongsearchcapability,butalsothestabilityhasbeeneffectivelyimproved.
简介:Basedonthenichegeneticalgorithm,theintelligentandoptimizingmodelfortherollingforcedistributioninhotstripmillswasputforward.Theresearchshowedthatthemodelhadmanyadvantagessuchasfastsearchingspeed,highcalculatingprecisionandsuitingforon-linecalculation.Agoodstripshapecouldbeachievedbyusingthemodelanditisappropriateandpracticableforrollingproducing.