Advanced Control Techniques for Batch Processes Based on Iterative Learning Control Methods

Korean Chem. Eng. Res., Vol. 44, No. 5, October, 006, pp. 45-434 { oh l om oh m -74 ne e (006 9o p r, 006 9o 3p }ˆ) Advanced Control Techniques for Batch Processes Based on Iterative Learning Control Methods Kwang Soon Lee Depart. of Chem. and Biomol. Engng., Sogang Univ.,, Shinsoo-dong, Mapo-gu, Seoul -74, Korea (Received September 006; accepted 3 September 006) k o lp r lp l r p v 0l m rl tep rl(apc) p p nr l p v p l. pl l t p s rl APC p pp kv p lvv k p. p p rp rrp o p q p p re v l q po p p. p rr p p APC p drl(ilc)l l r APC l pl p. l p p k er rl rn l nrp p, ILC p p r APC l p p e. h Abstract The operability and productivity of continuous processes, especially in petrochemical industries have made remarkable improvement during the past twenty years through advanced process control (APC) typified by model-based predictive control. On the other hand, APC have not been actively practiced in industrial batch processes typified by batch polymerization reactors. Perhaps the main cause for this has been the lack of reliable batch process APC techniques that can overcome the unique problems in industrial batch processes. Recently, some noteworthy progress is being made in this area. New high-performance batch process control techniques that can accommodate and also overcome the unique problems of industrial batch processes have been proposed on the basis of iterative learning control (ILC). In this review paper, recent advancement in the batch process APC techniques are presented, with a particular focus on the variations of the so called Q-ILC method, with the hope that they are widely practiced in different industrial batch processes and enhance their operations. Key words: Batch Process, Advanced Process Control, Iterative Learning Control. m rl(mpc, model-based predictive control) tep l rp rl v 0l krp r mp, p o ro rp trl p e q qk [, ]. pl p r q p p v rp n e MPC pp. pl l rp rl k rp n l pl. l l l po p. n r To whom correspondence should be addressed. E-mail: kslee@sogang.ac.kr l o p sr p p s q v p p l t pp v o l r~l v tp lr v kp, kr e p s r p vp n rl p pp p o k rp opp qn p. e rp l rp l r l n p p pl rl ql ke n pp v l r p. rp enrp r rl p o drl(ilc, iterative learning control) p p r l mp, l rl rp pn 45

46 p p p. l rp l rl p ILC p p rl p n pnl rp q m.. om Š rlp rl rp l r n l p. p p rl l opp p p enrp r rl p qp lr p. rp q t, r, ~, q, pk,, e k ll k ˆ pn pv, r lp rl l ~ l r p p ˆ. () l rp rl r n p o regulation r tlvv, rp rl r tracking r tlv. p rp tracking rl regulation p rp r l l servo p l. () rl p ˆ. p p rp r~l ~ r ˆl nr l ˆ p. l r rq~ rpv r r ˆ t l nr l p. rp p o p nrp v l p. p p rl e p r p l l. (3) rp rp r. rp p nr p erl vp k p. l r l r r ep p s r p rp l., sr p vrl ee l p vrp. (4) rp sq m rl p e l p. m l p p n er n sq p r k n r rl. (5) p rp v p ˆl nr l r p ee l n p op n v k. (6) rp pr ol p p rp r nrp. rl r p r pl l s erp r p n p pr v kp p. rp rl p ol p n p p plk. p p l m p r q e p p. k rl p rll p l m v m v l l er l l sp p pv v l l. 3. PIDoh i m oh q p r p l rl m p PID rl pn rl rp rl p. rp p e l v l p l rl spp o44 o5 006 0k v k rr rl p lp l n. er l~ t p p n, rl r q v kk r l rl p r n. Fig. p er t p l p m rl p p l p p. rl spp r l pv k k relay rl } q p. l t n p n t (ultimate period) p p. p p p t p p p p l n p p. p v p l p t m p p PIDrl sp p e l d u, Fig. m p rl lp pl. Fig., p p l m p r p rp l l qv k n rl p r l p lp p. v kp n, rl p l l. rl rlm v p r o qp r pr p ov pv k n rp rlm lk l. Fig. 3 p r ramp n rlm p l. er t p l r p p p m d er Œp l pp eq Fig.. Variation of the period of reactor temperature oscillation in an industrial polymerization reactor. Fig.. Improvement of temperature control in Fig. by sche-duling of PID parameters.

drl p rp rl 47 PID rl p r pr p ov r m n p l q rlm lk t p ppp q k r p. p r e p r o l r r p k n l. p e l r qp Fig. 3. By feedback-only action, PID control necessarily results in lagged response to ramp reference trajectory. erl p m rlp. rl pn n, rlm v erl pp p rp eq l pp, p erl p rl qp p p p m dp l n n. p r kp rlm v rl rl qp p. p rl q p p llv p p k, l t plk. p o e o r r p k p n. m l p e rp rp p t ut () R I ek ( ) ut () ut ( ) + R I et ( ) k 0. p qp rlm e(t) 0p k p kpp p lk o l u(t) v p eˆ p. R I l rlm e(t) 0p u(t) l pr p, k t v ˆ. p o R I r l p r ˆ n l. pr k rp pr k nr, r k r rp p Ž p tlv, n p Ž p p p dl r rl ()e p p f q rlm lk pp p. p p ep ˆ o l p p N er p, k w l p p, r rp p p rp q. () yt () Fut ( ()) () u k [ u k ( 0) T u k ( ) T u k ( N ) T ] T m p tlv r r(t)l r(t) y(t)p pn l u(t) p p e o e n p. rp ep m p l e p o e l p dv k rrp p. 4. om ooh i ol l rrl l rl p p / o rl p rl pn e t v l m. p e t p [3-6], p [7-0], r [-4], v r [5, 6], RTPm p ~ q [7-9] k rl e l m. p r l rl p p p r l ps, e p r p l o l np v l ll p p p. rp rn p n rp, o p p r p v n p p r k n p qœp lk. 5. ohm 5-. lm n oh q p rp rrl p m sq l l rl p p p. l l p p rln seˆ p l p p l l. p p po r rl p l p. y k [ y k ( ) T y k ( ) T y k ( N) T ] T r [ r ( )T r ( ) T rn ( ) T ] T p dl r rl k u k H ( r y k ) u k u k + He k n 0, p r rlm p po p H l e k 0p v k p. p p o l H r l l r n l. Fig. 4l o r qp rl p m. r rp Ž p pr, p Ž p n p op l H rr ˆ p l Fig. 4. Batch-wise integral control can completely remove the tracking error despite nontrivial model error. (3) (4) Korean Chem. Eng. Res., Vol. 44, No. 5, October, 006

48 p rlm tl mr rrl p p p. p rl k- p nr pn l k l n rlp e s v, pr l p n r dp rle rp l drl(ilc; iterative learning control) n p. u k k w p o e n p. ILC r p m l lim e k 0 seˆ r k o e pr p nr pn l rl p p. 5-. ohm i ILCl p p 35 r Garden[]l p kl pp, p Millerm Mallick[], Uchiyama [3]l p pn p. 984 Arimoto [4]l p s v qq p pv v ml. Arimoto p Žp o rp r s o l o D- ILC p rk m. p ILC rl l e p e k v, r l ps v k k v k v p k ˆ p rk l. l l ps v k k v rlm 0p o rl p p pr s p se k. p tlv n p s p seˆ p p d pp k rl l p k vp p p., p ILC SISO(single-input single-output) rp p mp l MIMO(multiple-input multiple-output) q l. p p p MIMO rl p rp e k 0p seˆ p ILC p rl e k minimum r lk. ILC r r k vp l l mp, n p s prp p r k v p l mp, p p u k u k +H e k +... +H N e k N (5) k vp rk m [5]. r r rl k vp prp tv v, r rl qpp pp, r ILC k vp p p k v m [6]. ol ILC ee rl p l l pp, k nrp v k, rp n l p v open-loop ˆl p. p rrp m o l ee rl p l p rk m. p rp ee rl p ILCl ILC p r v p m p. ILC p r kl rk p pn le kl q. pnl HDDm p pr r r, qlp q l p pnp p lv pv, v k 0,, ~, q kp s rl pnp v p. Ahn [7]l p s ILC l l o44 o5 006 0k p p p p. p rl rl p ILC p rp, k ILC t p e MIMO r ILC p p en p v prototype k v p. 6. ILCi ok oh 6-. om rp e rp v tlv r r t l nr nl p e (time-varying) rp p. rp l pv ILC m np pp p p l r m r v k. pr p p N erp p p p p e ˆ ep p. xt ( + ) At ()xt + Bt ()ut () + Kt ()vt () yt () Ct ()xt () + vt ()t, 0,,, N l v(t) rl op n rqpp op r p p innovationp p, rp zeromean,, erl p e p. kl p v(t) p v r pv kp p. (3)e p r l p e rp (6)ep p e pp y k Gu k d k l G g, 0 0 0 0 g, 0 g, 0 0........ g N, 0 g N, g NN, g n,m C(n)A(n-)...A(m+)B(m) (8) p, d k x(0), v(τ), 0 τ Np n e p. d k kl rp zero-mean e dˆ m d k p p p. d k dˆ + d k l d k p p qp np r r d k d k + n (6) (7) (9) (0) pr y k Gu k + d k, e k r y k, e k r y k rp, km k-l p (7)ep l r p p pp rpep l.

drl p rp rl 49 e k e k G u k + n e k + dˆ e k () l u k u k u k- p p. o l s p Lee [8]l q l p. 6-. Q-ILC ee rl l ILC r e k-, e k-,... u k r p. p p p p r l p r p. min -- { + u k R } e kk u Q k () l x Q x T Qx, k- vp nrr l l e kk lp e k p rm p p p Kalman el p p tlv. u k e kk e k k G u k e kk e kk + K( e k e kk ) (3) Fig. 5. Comparison of tracking performance of I-ILC and Q-ILC. rks p l n, ()ep p p tlv. u k He k k ( G T QG + R) G T Qe k k (4) p k vp Q-ILC(quadratic criterion-based ILC) Lee [8]l p rk p. Q-ILC sp ILCp rrp rp p p rp, enrp p prp v p. sp l l ILC u k e eˆ, p p nonsquare MIMO rl rn l p r p. r l rks p p n, (4)em p pep tlvv kv QP(quadatic programming) r l rp e p v. Q-ILC r G 00Í m v v rlm p eˆ p p p v. Fig. 5l SISO rl Q-ILCp p l l ILCm l m. l er r Q-ILC o r p G p () s 0.8 ( ----------------------------------- 5s+ ) ( 3s + ) G s., () m ( ----------------------------------- 6s+ ) ( s + ) (5) p t p r r m. r l qpp p p r m. 6-3. oh m Q-ILC ee rl(rfc, real-time feedback control) p l l p. ee rl p k l p p. l r RFCm ILC p el k vp. p k vp rp u k () t u k () t + H e k ( :t) + H e k ( :N)+ (6) p s. l H e k (:t) p RFCl, v p ILCl. pm p rl ep l l q ep (6)ep u k (t) p e k (t) ˆ ep eˆ p. p p p llv k r l p RFC ILC llv. pr r (6)ep el v(t) kl v p p qpp p dl r rp r. v k () t v k () t + nt () v k () t v k () t + vˆ () t l n(t)m vˆ (7) (t) rp qpp. pm p v(t)p l (6)ep n(t)m vˆ (t)p m p p l p. x k ( t+ ) At ()x k () t + B() u t k () t + Kt ()nt () y k () t y k () t + C()x t k () t + nt () xˆ k ( t+ ) At ()xˆ k () t + Kt ()vˆ () t ŷ k () t Ct ()xˆ k () t + vˆ () t y k () t y k () t + ŷ k () t (8) (9) (0) (8)ep ˆ ep o x k () t l y k ( ) p... y k ( N) q n ˆ rp lk. p n ˆp v y k () t y k ( t t) p y k k () t ˆ ep. p p, (8)e~(0)ep l p e k (t) p ˆ ep p. Korean Chem. Eng. Res., Vol. 44, No. 5, October, 006

430 p x k ( t+ ) xˆ k ( t+ ) At () 0 0 A() t + Kt () 0 0 K() t x k () t xˆ k () t nt () vˆ () t + Bt () 0 pep p p p rp q. u k () t e k () t e k ( tt) [ C() t Ct ()] x k() t n() t vˆ () t xˆ k () t z k ( t+ ) Φ()z t k () t + Γt () u k () t + Kt ()wt () ζ k () t e k () t e k ( tt) ()z t k () t + mt () () () pr RFC k vp r l k llv p. r p p LQG r p p. min --E u k N t 0 e k () t Q + u k () t R ζ k () t e k ( tt) + Q (3) u k (t) rlm p r p p d p r rl, v, ILC qp o p. e p r o l ˆ ()e u k (t) p p r p. (3)e rl r t LQG rrl r u k (t)p p l l p q re l p. p rl k vp Lee [9]p re mp BLQG(batch LQG) m. r l rks p p n LQG p l ov p p p MPC r p v. min -- u k p k Subject to linear constraints on process variables + (4) (4)el m p m e k (t+k t) o m ()ep pn l. rks p p n, (4)ep QP l rp llv. ˆ ep ˆ o mv (4)el p ILC rl Lee [30]p BMPC (batch MPC) m. l QILC, BLQG, BMPC r l ILC p Q-ILC l. p nl e k () t ζ k () t + e k ( tt) min p seˆ u k (t) ()ep el l rl RFC ILC. 6-4. m 6-4-. rrlm vrlp er r nrl q 3, 4 p vp n tn r qn. t p l v m rl p Q-ILC kl re k v p n p. o44 o5 006 0k u k () t R e k ( t+ k t) Q + u k ( t+ k ) R ζ k ( t+ k t) e k ( t+ k t+ k) + u k ( t+ k ) R + Q s r p vrl, nr t o p Œp n sq p, n r k k vl lk. r s r p vp v m p, p m p er vp r m l v rlp p p. p v m ep lp p m o v p rlp p r. ()el sr p vp ζ k (t)p n t rp q. p vp q erp ˆ z k (t)m er p v p u k (τ), t τ N p p p }k ()ep e p r lk. el BLQG BMPC rr k vp rn s vrl el Q-ILC. nr t r p s m p p n q pv, e e ˆ p. Chin [3]p Q-ILC ol l r p d p p v ppp mp rp p p el m. p l p p p rn ml pn p. 6-4-. ILCm RFCp rp ()ep, kl RFC k v (6)p p rp u k u k + H e k + H e k (5) m p p. p k vp RFCm ILC el v, v p v p. ILC u k r p o r rl p eˆ l qp v, ee l m kp n p e k l op u k rp l p r. p ILC k l np p e rp k. p rrp o l Chin [3]p p u k ILC o u k m RFC o û k u k u k + û k m p l p rp, p p l u k k vp rk TBC(Two-stage Batch Control) m. pm p p p l ()ep e k e k G u k + n + dˆ e k e k Gû k (6) m p. l u k p dˆ p m p v kp re p pn l u k H e k p, û k H e k u k (7) (8) m p ee n dˆ p m p. p ee n p m p v k p f ILCp r u k

l pˆ v k. p u k u k + û k lr rl r n. e p p k k vp p. e k vr r lp e k Kalman l r. 7. Q-ILC mk i 7-. m m j pžsoh Chin [3]l p p, p l p drl p rp rl 43 A+B C B+C D (9) pp v p. p nr p n Fig. 6l p. A l Œp 30 l B l rp Œp, 00 C o k(4 mol) p nrp rp. p rl Ap 0 p r p p Cp s p r n. Cp p s. rl p q m m B Œp p sq pp, p p m o rl ove ˆ el C p l p eˆ p rk. p rl v m rlm m r rl el BMPC q l l. Fig. 7, 8l rl me m. porp m p mp l v m r C p q r p p. p vrll Only Tracking Control p m rl p (q m sql p ) np, Only Inferential Control p v r p pn l ee rl Nominal Case inferential control s l p o rl(ilc) p. BMPC pn p p m rrl Žp p l e p l p v l [30]. 7-. RTPi lmœ j m oh RTP(rapid thermal processor) op p p n l ~ rp op annealing,, v k rp n q p. op l p p tungsten-halogen lamp p Fig. 6. Operation scenario of the semi-batch reactor example. Fig. 7. Batch-wise improvement of temperature profile of the semi-batch reactor example under BMPC with model uncertainty. Fig. 8. Batch-wise improvement of the production of C under BMPC with model uncertainty. Nominal case means measurement feedback at the end of each batch plus real-time inferential control. n l l, p op m p p ov rp s k. l l np Cho [33]l p p. Fig. 9 p 40 p Kw lamp v 8p op n RTP l e p mp, lamp p 0 grouping l s q n m. op l 8 p lr l p m rl m. rl ILCm RFCp p TBC n mp, TBC 8 0 rp r l p. Fig. 0l p 8r m p rrl p l l v lt p. Fig. p rp p r p m rp (y(t) T(t)) r l l TBC rn nm p o rn l r p r m p 4dp rp l(y(t)t 4 (t)) r p TBC, rn np p. p l p m p p Korean Chem. Eng. Res., Vol. 44, No. 5, October, 006

43 p Fig.. Performance comparison between T model-based TBC and T4 model-based TBC. Fig. 9. Thermocouple locations and the grouping of tungsten-halogen lamps for MV use. Fig. 0. Progressive improvement of tracking control performance of wafer temperatures under TBC. o44 o5 006 0k p p. Fig. 0 l temperature difference, temperature gapp 8vr rm pl q m. p rl p lp q l rn l n p. 8. oh e rp r ILCm p p rl p l ˆ p. ql p p. s np Leem Lee[6, 34]p. 8-. -D oh -D(-dimensional) rl -D p rl p. -D edšp ˆ rp p p d p pl. rp e p d p p ˆ rp p lv -D edšp m. rp e p d o l q l p -D edš s p v p. l, -D rl r rl np. q v -D edš rl l l m p k v o mlp p p v t l l m [35-38]. 8-. Run-to-run oh Run-to-run(RTR) rl r rp ~ rl p n l m rl p. RTRp e ILCp p, rp p n ( p ) l rr rp n, rp p e p k p nl ILC p. ~ rp n ee l op p ˆ r p n l l, nrs (p )p reˆ p ˆ ( ) r l RTR nrs p p p [6, 34, 39, 40]. RTR rl e rp r EVOP[4], batch-to-batch r [4, 43] vrp p p.

8-3. oh rp rp k l rp r rp pr t rp p. rl(rc, repetitive control)p pr t p nr l p t p rlp p p ILCm rp o, ILCp qrp p [44]. ILCm rp rp l rp ˆ eqrl reset r, pr t p ˆ p t p ˆl m p pr p k p. rp liftingp p p. Predictive Control Technology, Control Eng. Practice,, 733-764(003). 3. Song, I. and Rhee, H., Nonlinear Control of Polymerization Reactor by Wienr Predictive Controller based on One-Step Subspace Identification, Ind. Eng. Chem. Res., 43, 76-774(004). 4. Rantow, F. S., Soroush, M. and Grady, M. C., Optimal Control of a High-Temperature Semi-Batch Solution Polymerization Reactor, Proceedings of the 005ACC, Portland, USA, 30, June(005). 5. Sheibat-Othman, N. and Othman, S., Control of Emulsion Polymerization Reactor, Ind. Eng. Chem. Res., 45, 06-(006). 6. Cetinkaya, S., Zeybek, Z., Hapoglu, H. and Alpbaz, M., Optimal Temperature Control in a Batch Polymerization Reactor Using Fuzzy-Relational Models-Dynamics Matrix Control, Comp. Chem. Eng., 30, 35-33(006). 7. Hua, X., Rohani, S. and Jutan, A., Cascae Closed-Loop Optimization and Control of Batch Reactors, Chem. Eng. Sci., 59, 5695-5708(004). 8. Arpornwichanop, A., Kittisupakorn, P. and Mujtaba, I. M., Online Dynamic Optimization and Control Strategy for Improving the Performance of Batch Reactors, Chem. Eng. and Processing, 4, 0-4(005). 9. Bouhenchir, H., Cabassud, M. and Le Lann, M. V., Predictive Functional Control for the Temperature Control of a Chemical Batch Reactor, Comp. Chem. Eng., 30, 4-54(006). 0. Pieri, F., Caccavale, F., Iamarino, M. and Tufano, V., A Controller-Observer Scheme for Adaptive Control of Chemical batch Reactors, Proceedings of the 006 ACC, Minneapolis, USA, 554, June(006).. Ma, D. L, Tafti, L. KU and Braatz, R. D., Optimal Control and Simulation of Multidimensional Crystallization Processes, Comp. Chem. Eng., 6, 03-6(00).. Caliane, B. B. Costa, Aline C. da Costa and Rubens Maciel Filho, Mathematical Modeling and Optimal Control Strategy Development for an Adipic Acid Crystallization Process, Chem. Eng. and Processing, 44, 737-754(005). 3. Voller, U. and Raisch, J., Control of Batch Crystallization-A System Inversion Approach, Chem. Eng. and Processing, 45, 874-885(006). 4. Costa, C. B. B., da Costa, A. C. and Filho, R. M., Evaluation of Optimisation Techniques and Control Variable Formulations for a Batch Cooling Crystallization Process, Chem. Eng. Sci., 60, 53-53(005). 5. Vu, T. T. L., Durham, R. J., Hourigan, J. A. and Sleigh, R. W., Dynamic Modelling Optimization and Control of latose Crystallizations: Comparison of Process Alternatives, Sep. and Purif. Technology, 48, 59-66(006). 6. Ronia M. Oisiovici* and Sandra L. Cruz, Inferential Control of High-Purity Multicomponent Batch Distillation Columns using an Extnded Kalman Filter, Ind. Eng., Chem. Res., 40, 68-639 (00). 7. Ulas, S., Diwekar, U. M. and Stadtherr, M. A., Uncertainties in Parameter Estimation and Optimal Control in Batch Distillation, Comp. Chem. Eng., 9, 805-84(005). 8. Da, L., Kumar, V. G., Tay, A., Al Mamun, A., Weng Khuen Ho, See, A. and Chan, L., Run-to-run Process Control for Chemical Mechanical Polishing in Semiconductor Manufacturing, Proceedx k + ( 0) Ax k ( 0) + Pu k y k x k ( 0) + Gu k + d k drl p rp rl 433 (30) e l ˆ reset (x k (0)0), p p (7)e p v. RC p l (3), (4)e p r l rl. Lee [45]p MPC r l p RC rk p RMPC(Repetitive MPC) m. 9. rp rl p lp p rl p pp, o rl el rl p p. rp r r p l p dv k l, r rp o rl p rrs p seˆ l r v p. ILC p rrp rp p p, v l ~ ILC p rp rl p krp rp m. p rl p r p rrp n, l rn MPCp qrp rl pl t pn m l. l ILC p rrl p rp, enrp p v p s m. l r v r rl l p lp p p p n l kv v rr rl p re l pv kk pnp n r rp pl p ep. ILC p rl p rrlp n qp ll p. p q r l r R0-006-000-377-0 vop p pl. y. Morari, M. and Lee, J. H., Model Predictive Control: Past, Present, and Future, Comp. and Chem. Eng., 3(4-5), 667-68 (996).. Qin, S. J. and Badgwell, T. A., A Survey of Industrial Model Korean Chem. Eng. Res., Vol. 44, No. 5, October, 006

434 p ings of the 00 IEEE Int. Sym. on Intelligent Control, 740-745, Vancouver, Canada(00). 9. Emami-Naeini, A., Ebert, J. L., Kosut, R. L., de Roover, D. and Ghosal, S., Model-based Control for Semiconductor and Advanced Materials Processing: an Overview, Proceedings of the 004ACC, 5, 390-3909, Boston, USA(004). 0. Matsumoto, K., Suzuki, K., Kunimatsu, S. and Fujii, T., Temperature Control of Wafer in Semiconductor Manufacturing Systems by MR-ILQ Design Method, Proceedings of 004 IEEE Int. Conf. on Control Appl.,, 40-44, Taipei, Taiwan(004).. Garden, M., Learning Control of Actuators in Control System, UUSU Patent No. 3,555,5(97).. Miller, R. C. and Jr. Mallick, G. T., Method of Controlling an Automatic Machine Tool, U.S. Patent No. 4,088,899(978). 3. Uchiyama, M., Formation of High Speed Motion Pattern of Mechanical Arm by Trial, Trans. of the Society of Instrum. and Control Eng., 9, 706-7(978). 4. Arimoto, S., Kawamura, S. and Miyazaki, F., Bettering Operation of Robotics by Learning, J. of Robotic Systems, (), 3-40(984). 5. Bien, Z. and Huh, K. M., Higher-Order Iterative Learning Control Algorithm, IEE Proc. Part D on Control Theory and Appl., 36, 05-(989). 6. Lee, K. S. and Lee, J. H., Iterative Learning Control-based Batch Process Control Technique for Integrated Control of End Product Properties and Transient Profiles of Process Variables, J. Process Control, 3(7), 607-6(003). 7. Ahn, H., Chen, Y. and Moore, K., Iterative Learning Control: Brief Syrvey and Catgorization 998-004, IEEE Trans. Systems, Man and Cybernetics-Part C: Applications and Reviews, Accepted for Publication(006). 8. Lee, J. H., Lee, K. S. and Kim, W. C., Model-based Iterative Learning Control with a Quadratic Criterion for Time-varying Linear Systems, Automatica, 36(5), 64-657(000). 9. Lee, K. S., Lee, J., Chin, I. and Lee, J. H., Control of Wafer Temperature Uniformity in Rapid Thermal Processing Using an Optimal Iterative Learning Control Technique, Ind. Eng. Chem. Res., 40(7), 66-67(00). 30. Lee, K. S., Lee, J. H., Chin, I. and Lee, H. J., A Model Predictive Control Technique Combined with Iterative Learning for Batch Processes, AIChE J., 45(0), 75-87(999). 3. Chin, I. S., Lee, K. S. and Lee, J. H., A Technique for Integrated Quality Control, Profile Control, and Constraint Handling for Batch Processes, Ind. Eng. Chem. Res., 39(3), 693-705(000). 3. Chin, I., Qin, S. J., Lee, K. S. and Cho, M., A Two-Stage ILC Technique Combined with Real-Time Feedback for Independent Disturbance Rejection, Automatica, 40(), 93-90(004). 33. Cho, M., Lee, Y., Joo, S. and Lee, K. S., Semi-empirical Modelbased Multivariable Iterative Learning Control of an RTP System, IEEE Trans. On Semicon. Manuf., 8(3), 430-439(005). 34. Lee, J. H. and Lee, K. S., Iterative Learning Control Applied to Batch Processes: An Overview, accepted for publication in Control Eng. Practice(006). 35. Roesser, R., A Discrete State Space Model for Linear Image Processing, IEEE Trans. A.C., 0, -0(975). 36. Kaczorek, T., Two-Dimensional Linear Systems, Springer-Verlag, Berlin(985). 37. Kurek, J. E. and Zaremba, M. B., Iterative Learning Control Synthesis based on -D System Theory, IEEE Trans. A.C., 38, -5(993). 38. Shi, J., Gao, F. and Wu, T-J., From Two-Dimensional Linear Quadratic Optimal Control to Iterative Learning Control. Paper. Iterative Learning Control for Batch Processes, Ind. Eng. Chem. Res., 45, 467-468(006). 39. Chen, Y.-H., Su, A.-J., Shiu, S.-J., Yu, C.-C. and Shen, S.-H, Batch Sequencing for Run-to-Run Control: Application to Chemical Mechanical Polishing, Ind. Eng. Chem., Res., 44, 4676-4686(005). 40. Firth, S. K., Campbell, W. J., Toprac, A. and Edgar, T. F., Justin-time Adaptive Disturbance Estimation for Run-to-Run Control of Semiconductor Processes, IEEE Trans. On Semicon. Manuf., 9(3), 98-35(006). 4. Wilde, D. J. and Beightler, C. S., Foundations of Optimization, Prentice-Hall, Englewood-Cliffs(967). 4. Zafiriou, E. and Zhu, J. M., Optimal Control of Semi-Batch Processes in the Presence of Modeling Error, Proc. of 990 ACC., San Diego, USA, 644-649(990). 43. Zafiriou, E., Adomaitis, R. A. and Gattu, G., Approach to Runto-Run Control for Rapid Thermal Processing, Proc. of 995 ACC., Seattle, USA, 86-88(995). 44. Hara, S., Yamamoto, Y., Omata, T. and Nakano, N., Repetitive Control System: a New Type Servo System for Periodic Exogeneous Signals, IEEE Trans. A.C., 33, 659-668(998). 45. Lee, J. H., Natarajan, S. and Lee, K. S., A Model-based Predictive Control Approach to Repetitive Control of Continuous Processes with Periodic Operations, J. Process Control,, 95-07(00). o44 o5 006 0k