:2007-12-21:宋保维(1963- ),男,辽宁葫芦岛人,教授、博导,主要研究方向为水下航行器设计,水下特种减阻降噪技术等。:1002-0640-(2008)-0133-03iSIGHT宋保维,李 楠(,710072):,,,,iSIGHT,iSIGHT,iSIGHT:iSIGHT,,,Pareto:TP183:AApplicationResearchofiSIGHTinSolvingMulti-objectiveDesignOptimizationProblemsSONGBao-wei,LINan(CollegeofMarine,NorthwesternPolytechnicUniversity,Xi'an710072,China)Abstract:Inengineeringareas,singledesigntargethasnotmadedesignerscontentwithincreasingofdesigndemands,somulti-objectivedesignoptimizationproblemshavegraduallydominatedindesignareas.Nowthesearesomanymethodsofmulti-objectivedesignoptimization,andtheyallhavetheirownadvantagesandshortcomings.ThisthesismainlyintroducesiSIGHTwhichisakindofDesignOptimizationFramework.MeanwhilewewillresearchapplicationcircumstancesofiSIGHTinthemulti-objectivedesignoptimizationproblemofshellstructureoftorpedo.AndwewoulddiscussfeasibilityofiSIGHTinsolvingmulti-objectivedesignoptimizationproblems.Keywords:ISIGHT,multi-objectivedesignoptimization,shellstructureoftorpedo,paretosolutions引 言(multi-objectivedesignoptimization),,,,,,,,(Pareto)Pareto[1],,,,,iSIGHTiSIGHT,Daratech,iSIGHT,iSIGHTParetoEngineeringDataMining,iSIGHT1 多目标优化问题:Vol.33,SupplementJune,2008FireControlandCommandControl33Minimizef(x)=fm(x)m=1,2,,Ms.t.gj(x)0j=1,2,,Jhk(x)=0k=1,2,,Kx(L)ixix(H)ii=1,2,,I,Minimize,f(x)fm(x)Pareto,,;Pareto,,,,Pareto,,1985,J.D.Schaffer,VEGA;1993,C.M.FonsecaP.J.FlemingParetoOptimalityMOGA;1994,N.SrinivasK.DebMOGA,NSGA;1999,E.ZitzlerL.Thiele(Elitism),SPEA,;2001,E.ZitzlerSPEA,SPEA,SPEA,NCGA;,K.DebS.AgrawalSPEA,NSGA,NSGA,,Pareto,iSIGHTNCGANSGA2,,,,ParetoiSIGHTmonitorToolsEngineeringDataMining()Pareto,,,2 相关理论概述[3],,,,,,,,,,,:;D=0.5334m,p0=1.275MPa2.1pjn1=1.23,pj=n1p0=1.568MPa(1)2.2,20A,s=215.75MPa,E=1.96105MPa,=0.3,:t1.05pjR0.85s=0.2399cm(2)t,[0.2399,+]2.3ll1.029pj(100tR)32100t+0.62Rt=14.98cm(3)l(-,14.98cm]2.41)02U=0.6425lRt,=4.337,K02lt,02=K02pjRt0.85s=183.38MPa(4)2)rKrltr=KrpjRt0.55s=118.66MPa(5)3)(leq)x=12K1lt(leq)x=12=(0.91K1-0.3Kr)pjRts=215.75MPa(6),2.51)pcrU=0.6425lRt,=Rl,=1.92,n=pcr=En2+0.52-1[t3(n2+2-1)212(1-u2)R3+t4R(n2+2)2](7)2)pcr1=0.75,2:cr=K021pcrRt,=crs2=0.01271184-0.06355893+0.1906832-0.625827+1.29479cr=2cr0.85s(8)pcr=12pcrpj(9)cr,pcr2.61)I,2mm1342008I=(AiZ2i+I0i)+(AiZi)2Ai(11)2)(pcr)gL=400cm,1=RL=0.2,n1=2(pcr)g=E1+212(n21-1)[tR41(n21+21)2(n2-1)+(n2-1)IR3l](12)3)(pcr)gcr=(Kr+0.15)1(pcr)gRt,=crs,1=0.752=0.01271184-0.06355893+0.1906832-0.625827+1.29479(pcr)g=12(pcr)g(13)(pcr)g,(pcr)g(1.11.3)pj2.720A,=7.81000kg/m3,Ar,R1:m1=2Rt;:m2=2R1Ar;:m=m1+Llm2m33 iSIGHT在解决鱼雷壳体结构中多目标设计优化问题中的应用3.1:Min(m);Min[-(pcr)];Min[-(pcr)g];Mincrs.t.t0.2399cm0cml14.98cm02183.38MPa;r118.66MPa;(leq)x=12215.75MPa;cr183.38MPapcr1.568MPa;(pcr)g1.568MPa:t;l;l1l2l3l4(:Z,l1l2l3l4Z)64ParetoPareto,iSIGHT3.2iSIGHT3.2.1 在iSIGHT中对优化问题的描述iSIGHT,1,6(),4(Min)13()13.2.2 优化算法的选择,,iSIGHTNCGANSGANCGANSGA,,NCGANSGA,NCGA3.2.3 Pareto解的后处理Pareto,iSIGHTEDM(Pareto)64Pareto2,323Pareto,2DesignVariables,Responses(Pareto)3Pareto3.2.4 结果分析Pareto,t:l:l1:l2:l3:l41:1:1:1:1:1Pareto:m=37.56kg/m3;pcr=1.573MPa;(pcr)g=2.231MPa;cr=183.21MPat=0.2448cm;l=14.0457cm;l1=0.2103cm;l2=1.2245cm;l3=2.1340cm;l4=0.7361cm[2],m=38.33kg/m3;pcr=1.569MPa;(pcr)g=2.089MPa;cr=183.43MPat=0.24cm;l=14.98cm;l1=0.2cm;l2=1.2cm;l3=2.2cm;l4=9.5cmISIGHTpcr(pcr)g,;cr;m,,ISIGHT(下转第157页)135,:iSIGHT,,(9),,,,,9364,(10),103,:73,53,:31,,,3 结 论,,,,,,,31;;,,,:[1].[M].:,2003.[2].[J].,2003(2):32-35.[3].[J].,2005(2):10-13.[4].[J].,2006(9):7-8.[5].[J].,2004,16(5):56-61.[6].[J].,2005.[7].BP[J].,2007,24(2):15-17.[8].[J].,2004,16(2):29-41.[9].[J].,2007(4):L2.[10].[M].:,2000.[11],.[M].:,2001.[12].[M].:,1999.[13].[R].:,2003.[14].[J].,2005.[15]ForresterJW.IndustrialDynamics.Cambridge[M].MA:MITPress,1961.[16]ForresterJW.UrbanDynamics.Cambridge[M].MA:MITPress,1969.[17]ForresterJW.WorldDynamics.Cambridge[M].MA:Wright-AllenPress,1971.[18]EberleinRL.SimplificationandUnderstandingofModels[J].SystemDynamicsReview,1989,6(1):.[19]ForresterJW.LessonsFromSystemDynamicsModeling[J].SystemDynamicsReview,1987,3(2):136-149.(上接第135页)4 总 结,,,,,,,,,,iSIGHT,ISIGHTParetoEDM,,ISIGHT,ParetoPareto,,ISIGHT:[1].[D].:,2006.[2].[M].:,1998.[3].[M].:0,1997.[4]ShinyaW,TomoyukiH.Multi-ObjectiveRectangularPackingProblemandItsApplications[C].EMO'03,2003:565-577.[5]ShinyaW,TomoyukiH.NeighborhoodCultivationGeneticAlgorithmforMulti-ObjectionOptimizationProblems[C].SEAL-2002,2002:198-202.157,:SD