基于神经网络的港口运输优化方法

()Rn+(n),,,NEURALNETWORKBASEDPORTTRANSPORTATIONOPTIMIZATIONMETHODLiZhiYaoZhubinZhangWangxing(TheIndustrialPort)AbstractThemostreasonablelinefortheporttransportationhasbeenoptimallycalculatedbywayofRn+neuralnetworkbasedalgroithm.Emulationresultsdemonstratethattheneuralnetworkbasedoptimalalgorithmiscorrectandapplicabletothemathematicprogrammeintheotherfields,thusopeningupanewwaytotheoptimalsolution.Keywordsneuralnetworktransportationoptimizationemulation,,,1,Lagrange,,,Lagrange,,G.Dantzig1947,,,1986,TankHopfieldTH(1)()Lya2punov,,VLSI,,1TH2,()8238120001WISCOTECHNOLOGYVol.38No.1Jan.2000,,,,121tB1B2B3B4B5B6A1500A2700A3800A4600A5900A6100075080040090085080045002/tB1B2B3B4B5B6A17.06.08.05.56.27.4A28.25.24.55.78.34.0A39.06.88.06.24.75.9A45.06.27.17.75.85.2A55.18.36.55.27.57.0A66.88.54.56.17.86.0:minK=ijCijxij(i,j=6)(1)S.t.ixij=aj(i=6)jxij=bi(i=6)xij0(i,j=6)K/i=1,2,,6j=12,6Cijij/t-1ajj/tbii/txijij/t,,xij,xij,,,,(Lya2punov),,,,(),,Rn+Rn+:dx/dt=Tx+Ix0(2)xRn+={xRn|x0}TnnIRn(2)Rn+,,1(1)PxoRn+,(2)x0,(2)(2)[0,]1(2)x=x3,(1)(2)t,x3,x3Rn+2(2)E(x)=-12xTTx-xTI(3):(3)E(x)92:dE/dt=dE/dxdx/dt=-(Tx+I)dx/dt=-(dx/dt)20dx/dt=0,dE/dt=0,(1)E(x)(2)dx/dt=0,dE/dt=0,Rn+,E(x),E(x)Rn+xminRn+2,66=36,,12:xi=g(ui)=uiui00i0:yi=f(zi)=2xi(b1,b2,b6,a1,a6)T,(c1,c2,c36)T,2,z,-AT=(-aij)T1236,A(,qi=1,Ri=12)KCL,dx/dt=-C-2x-2AT(Ax-b)x0(4)(4),xmin312(4),=0.01,=100,3,,(52)03200011.525.4mm,7002100mm(2230mm),0.355.0mm,9002050mm,,4mm,,(2),TRT(3),,WTO,,4,,,,10,;,,,,,,1.,,1998,(6):352..,1999,(2):1163..,1999(1):46474..,1999,(1):275.WTO.,1999,(17):1136.1999.,1999,(7):15227.1998,1738.21.,1999,(2):16249..,1999,(6):11510..,1999,(5):11011..,1999,(5):111812..,1994,(19):1324(:1999-08-13)(30)3B1B2B3B4B5B6A1500500A2160540700A3800800A441014050600A5340560900A640034026010007508004009008508004500:minK=2.034()410,AI,,,,,1..:,19952..:,19963..:,1984(:1999-08-05)2520001