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




гʼĵѧϰ
Ʊѧ
ؼʡʼѧϰƣ.
ITERATIVE LEARNING CONTROL ALGORITHMS WITH INITIAL
UPDATE ACTION
HUANG BaojianSUN MingxuanZHANG Xuezhi
(Department of Electronics,Xi'an Institute of Technology, Xi'an 710032)
Key wordsInitial condition, iterative learning control, convergence.
1
õѧϰ㷨ƿʱΪ˱֤㷨ԣϵͳĳ̬޶һνĳʼ.Ŀǰ״Ҫ̬ϸظ̬15.Ȼʵʵظλ̬̬ƫ.ں̬ܶδ֪ģϵͳ̬Ҳǹ̶.оڵ̶̬µѧϰ⣬˴гʼĵѧϰ㷨㷨ԣ㷨ļ޹켣.
2Ҫ
һϵͳ
(1)
tʣ0,T,x,f(x)Rn,B(x)Rnr,y,g(x)Rm,uRr.ڸϵͳ¼ٶg(x)xĵgx(x)ڣf(x)gx(x)B(x)xȫһLipschitzx1,x2Rn,f(x1)-f(x2)kfx1-x2,h(x1)-h(x2)khx1-x2,hʣB,gx;B(x),gx(x)н磺B(x)M1,gx(x)M2.
Ըϵͳʽѧϰɣ
(2)
tʣ0,T,ek(t)=yd(t)-yk(t)(y(t))Ϊѧϰ㦣(y(t))ܦã
(3)
Ϊʼ.
1.ʽ(1)ϵͳѧϰ(2)
c1)I-(g(x))gx(x)B(x)ܦ<1,xRn;
c2)xk(0)=x0,k=0,1,2
kʱyk(t)ڣ0,Tϵļ޹켣Ϊ
(4)
kʱyk(t)ڣh,Tһyd(t).
֤. e*k(t)=y*d(t)-yk(t).ʽ(4)֪y*d(0)=g(x0)ˣe*k(0)=y*d(0)-yk(0)=0,y*d(h)=yd(h),.ȡһu*d(t)ʹϵͳ̬λx0켣Ϊy*d(t).Ǧu*k=u*d-uk,e*k=y*d-yk.ʽ(2)0t<hʱ
(5)
htTʱy*d(t)=yd(t)e*k(t)=ek(t)h(t)=0ʽ(5)Ҳʽ(5)κtʣ0,Tݾ.ɴ˿֤1Ľ.
3ᡡ
ѧϰ㷨ڴڳ̬ƫƵѧϰϵͳ1˶Ե̬Ҫ.h(t)ʵh,T϶켣ȫ٣ѧϰ㷨ڿ˷̬ƫƵӰʮЧģhСϴĿ.еĽ˵̬ϸظʱйؽ.
߼飺ƱУ196112.˶ʿ.о򣺼κۣѧϰ.
У19616.˶ʿ.о򣺵ѧϰƣſƣ.
ߵλҵѧԺϵ710032
ο
1Ahn H S, Choi C H. Iterative learning control for linear systems with a periodic disturbance. Electronics Letters, 1990,26(18):15421554
2Arimoto S, Kawamura S, Miyazaki F. Bettering operation of dynamic systems by learning: A new control theory for servomechanism or mechatronics systems. In:Proc.of 23rd IEEE Conf. on Decision and Control. Las Vegas, NV, 1984,10641069.
3Ortega J M, Rheinboldt W C. Iterative Solution of Nonlinear Equations in Severial Variables. Academic Press, 1970
4Poter B, Mohamed S S. Iterative learning control of partially irregular mutivariable plants with initial state shifting. Int.J.Systems Sci., 1991, 22(2):229235
5Sugie T, Ono T. An iterative learning control law for dynamic systems. Automatica, 1991, 27(4):729732
ոڣ1997-03-28
޸ڣ1998-09-28
