Զѧ
ACTA AUTOMATICA SINICA
1999ꡡ255ڡVol.25 No.5 1999




ϵͳԤƵĴԷ1)
ϯԣ
ժҪһϵͳоնԼȶԤƲԴͳſƵĴ.ͨԤƵʱŻʣõԤƴԵϽ磬ҽýӦϵͳõһĴָ.
ؼʡʱƣԤƣſƣԣϵͳ.
THE SUBOPTIMALITY ANALYSIS OF PREDICTIVE CONTROL
FOR CONTINUOUS NONLINEAR SYSTEMS
XI YugengGENG Xiaojun
(Institute of Automation,Shanghai Jiao Tong University,Shanghai200030)
AbstractConsidering the general continuous nonlinear systems,the suboptimality of stable predictive control with terminal constraints is studied and compared with the traditional optimal control.By analyzing the property of receding finite-horizon optimization,the upper bound of suboptimality is obtained.The result is applied to linear systems,and a suboptimality quantized index is proposed.
Key wordsReceding horizon control,predictive control,optimal control, suboptimality, continuous nonlinear system.
1
ʱRHC(Receding Horizon Control)ΪԤƵһʽʱŻܹʵɣһֱܵƽĹע.ʮKwonPearson12ݾͶʱϵͳRHCо֤նԼRHCϵͳȶԣRiccati̽ƳRHCſƵָ֮ĹϵRHCĴ.Mayneˣ34ݽնԼϵͳоʱϵͳRHCԣõȶԵĳûз.ڴ˻ϷϵͳRHCʱſƵĴԣһӦϵͳõ˸ĽۣKwon1ݵĽ.
2
ϵͳ
(1)
x(0)֪x(t)Rn,u(t)Rmf(0,0)=0.f(,)ģҹxLipschitz.ֱϵͳſƲԺնԼRHCԣǾʹϵͳȶ.
1)ʱſƲ
ϵͳ(1)ʱſu(t):0,)RmΪŽ⣺
(2)
LL(0,0)=0,ҴһǼ:0,)0,),(0)=0ʹö(x,u)(0,0),0ܦ(x,u)L(x,u).(..)ΪRnRmϵķ).Ȼ֪L(x,u)=0ҽ(x,u)=(0,0).
2)նԼRHC
tʱ̣֪ϵͳ״̬x(t),ʱŻ⣺

(3)
õſƽΪv*(;x(t),t),0ܦӡT.ʱʵʩϵͳĿΪ
(4)
Ȼtʱ̵ָJ*1(t)ȡx(t)ԱʾΪRnRӳ
J*1(t)=g(x(t)).(5)
ԿRHCõĿɺſɲͬȻÿһʱ̵Ŀʱ(3)Ž⣬ȫָ(2)ȴŵ.
3ϵͳRHCĴ
Уָʽ(2)LΪʽ(5)еJ*1(t)΢Ĵʤ£
1.һʽϵͳ(1)նԼRHCϵͳӦʱſָ(2)Ͻ磬ֵΪһŻʱָ.
֤.ǰRHCϵͳУtʱ̵ſΪv*(,x(t),t),0ܦӡTӦ״̬켣Ϊz*(),0ܦӡTӦָΪ
J*1(t)=L(z*(),v*())d,z*(0)=x(t),z*(T)=0.
t+tʱɣ

ÿڦӡʣ0,T-tʱڼΪtʱ̿ɵz(T-t)=0ڦӡʣT-t,Tʱu()=0fʿ֪z()=0,ӡʣT-t,Tݣرz(T)=0ʾÿt+tʱ̵նԼΪʱ̵ĿпƣӦָΪ

t+tʱָΪJ*1(t+t)

(6)
ʽ(6)Ҷ˵x(t),u(t)ֱӦRHCϵͳtʱʵ״̬Ϳ˸ݼ΢˼޴.ʽ(6)˷ֱt=0t=޻֣õ

ʽΪRHCϵͳӦſָʽ(2)ȫָΪJ*10J*1(0)Ϊʼʱʱָ.RHCȶԣ3ݣJ*1()=0
(7)
ע.ʵʿϵͳƳҪջϵͳȶҪջϵͳĳЩָҪ.ף4ԲȷϵͳνɱɣʹջָJxT0Px0,x0ΪϵͳֵPΪһ󣬶ʽͱĵóԤȫָĴԹϵ.ǶҲΪԤʵʵһ״̬ɱƣʹջɱֵҲȫָ겻ĳȷĽ.
ϵͳ(1)ſָ(2)ΪJ*0ʽ(7)Ӧ
J*0J*10J*1(0).(8)
۸J*10J*0Ĺϵ.
1.նԼRHCĴԺŻʱTѡȡйأTԽԽ.
2.նԼRHCȫŻгֵıƽ.
۲֤.ǱŻʱTնԼRHCɿɳֽӽſƣ1Ͻӽ̶ȵĶӳͨTõӦϽ磬ݴRHCϵͳ.
1ӦϵͳԵõĽ.
3.ϵͳնԼRHCԣӦſƵָϽ磬
J*10xT0P(0)x0,
P(0)ΪRiccati
(9)
ֽ.
ϵͳնԼRHCĴԽKwon1977Riccati̽ƳĽһµģ1ݣߵķֻϵͳ.ˣĵĶ1и㷺ĸ.
Riccati̽ĵԣŻʱTԽ󣬻Riccati(9)õP(0)ֵԽСTڡʱP(0)ֵݼΪRiccati̵ĽP.ˣϵͳնԼRHCԵĴԿñֵ
J0(x0)/J*0(x0)ܡx02P(0)/x02P
.ֵԽСԽ.
4
ſõĿƲȫŵģԤʱĹŻõĿȫֵſǴŵ.ڴһʽϵͳʱƴԵϽ磬óŻʱʹԵĹϵ.ȻӦϵͳõ۴Եʽ.ЩԤƵۺʵһ.
1) ȻѧĿ(69774004).
߼飺ϯԣ19461968ҵڹ¹ѧԺ1984ڵ¹ѧʿѧλ.ΪϺͨѧԶϵڡʿʦ.С̬ϵͳۡԤơ.Ҫ¸ϵͳ۵о.ĿǰҪоǸӹҵ̵ŻƼܻ˿.
1972.1996ҵѧ˶ʿѧλΪϺͨѧԶϵʿо.Ҫ·ԤƵо.
ߵλϺͨѧԶоϺ200030
ο
1Kwon W H,Pearson A E.A modified quadratic cost problem and feedback stabilization of a linear system.IEEE Trans.Automat.Contr.,1977,22(5):838842
2Kwon W H,Pearson A E.On feedback stabilization of time-varying discrete linear systems.IEEE Trans.Automat.Contr.,1978,23(3):479481
3Mayne D Q,Michalska H.Receding horizon control of nonlinear systems.IEEE Trans.on Automa.Contr.,1990,35(7):814824
4 Petersen I R,Mcfarlane D C.Optimal guaranteed cost control and filtering for uncertain linear systems.IEEE Trans.on Automa.Contr.,1994,39(9):19711977
ոڣ1998-02-16
޸ڣ1998-05-13
