可逆冷连轧卷取张力系统中的预测控制应用

254Vol.25No.420038JournalofUniversityofScienceandTechnologyBeijingAug.20032002–10–2329No.NERCAR-28()AGC––(DMC)11.1[1]1.GLT(s)PIGsPI.h=511.2[2,3]Gs=TN=DEwh60iL1s+(1)DEwhiL=V/LVV=D60iN,N1Gs,100083–PI.TG334.9;TG331;TP273Vol.25No.4•355•2(DMC)2[4,5]33.1{ai}MATLABSIMULINK,3.T=0.02s.3t0.4sN=2020{ai}:20×1=[0.070.140.210.280.350.420.490.560.630.70.770.840.910.981.051.121.191.261.331.4]T.3.2{ai}M=4P=20P×M==0=+1(2)=[–0.000014.2857–4.3956–3.8462–3.2967–2.7473–2.1978–1.6484–1.0989–0.549500.54951.09891.64842.1978]h1=1hi=0.5,i=2,,20=[10.50.50.5]T.4.hy0*uuDMCy*yk+1ek+1y0*15×4=0.070.140.210.280.350.420.490.560.630.700.770.840.910.981.0500.070.140.210.280.350.420.490.560.630.700.770.840.910.98000.070.140.210.280.350.420.490.560.630.700.770.840.910000.070.140.210.280.350.420.490.560.630.700.770.842Fig.2Predictivecontrolofthecoilingtensioncontrolsys-temT*+ekNtukgg100200300400500600/ms3Fig.3Staircasecurveofthecoilingtensioncontrolsystem00.40.81.21.61Fig.1TensioncontrolsystemmodelGsT+1.5s+0.3U,i+0.926s+61.73355U+2.128I0.47E1n0.1095TGsSCRss0.0017s+10.015s+10.0774s0.6066s+0346IGs0.01870.002s+1•356•20034.PIDDMC53.3KTK=0.1095T=0.346T0.3460.8PIDMC6K0.10953PIDMC7u80200400600800200300400500600/ms/ms6T(a)(b)Fig.6Curvesof(a)thePIDand(b)thepredictiveDMCarithmeticcontrolsystemwhenTischanged00.80.41.200.80.41.2(a)(b)200400600800200300400500600/ms/ms7K(a)(b)Fig.7Curvesof(a)thePIDand(b)thepredictiveDMCarithmeticcontrolsystemwhenKischanged00.80.41.200.80.41.2(a)(b)4Fig.4Flowchartofthedynamicsmatrixcontrol0100200300400/ms5DMCFig.5StaircasecurveofthepredictiveDMCarithmeticcontrolsystem00.40.81.2Vol.25No.4•357•4PIDMC1.[M].1996.632,.[J].1998(5):133.[J].2000(3)2334[M].19935,.[J].1999(4)106.MATLAB[M].1997.607[M].,1995.277ApplicationofPredictiveControltoaCoilingTensionControlSystemLIJing,WANGJingNationalEngineeringResearchCenterofAdvancedRollingTechnology,UniversityofScienceandTechnologyBeijing,Beijing100083,ChinaABSTRACTThetensioncontrolofsomedoublestandreversingcoldrollingmillwasaccomplished.Thetheoryandimplementalmethodswerediscussed.Thedynamicsmatrixcontrolwasappliedinordertoresolvethefluctua-tionofthetensioncontrolsystem.Simulationresultsindicatethatthedynamicsmatrixcontrolmaintainsthetensioncontrolprecision.KEYWORDScoillingtension;dynamicmodel;predictivecontrol;dynamicsmatrixcontrolcoldrollingmill01002003004005000100200300400500/ms/ms8PID(a)(b)uFig.8Curvesof(a)theincrementcontrolofthePIDarithmeticcontrolsystemand(b)theincrementcontrolofthepredic-tiveDMCarithmeticcontrolsystem–0.400.40.800.20.40.60.8(a)(b)