Conference PaperPDF Available

A Fault-tolerant Algorithm with Cycled Resetting Discount Factor in Semiconductor Manufacturing Industry

Authors:

Abstract

The threaded-EWMA run-to-run control is an important stable control scheme. However, the process outputs will deviate largely in the first few runs of each cycle if the disturbance follows an IMA(1,1) series with deterministic linear drift/fault and the thread has a long break length. In this paper, we analyzed the output of the threaded-EWMA run-to-run control. Based on the analysis of system performance, cycled resetting (CR) algorithm for discount factor is proposed to reduce the large deviations as well as a step fault to achieve the minimum asymptotic variance control. By analysis the influence of the fault, discount factor resetting and fault tolerant (RFT) approach is presented. Simulation study showed that the proposed approach is effective.
$EVWUDFW² 7KH WKUHDGHG(:0$ UXQWRUXQ FRQWURO LV DQ
LPSRUWDQW VWDEOH FRQWURO VFKHPH +RZHYHU WKH SURFHVV RXWSXWV
ZLOO GHYLDWH ODUJHO\ LQ WKH ILUVW IHZ UXQV RI HDFK F\FOH LI WKH
GLVWXUEDQFHIROORZVDQ,0$ VHULHVZLWK GHWHUPLQLVWLFOLQHDU
GULIWIDXOWDQGWKHWKUHDGKDVDORQJEUHDN OHQJWK,QWKLVSDSHU
ZH DQDO\]HG WKH RXWSXW RI WKH WKUHDGHG(:0$ UXQWRUXQ
FRQWURO %DVHG RQ WKH DQDO\VLV RI V\VWHP SHUIRUPDQFH F\FOHG
UHVHWWLQJ &5 DOJRULWKP IRU GLVFRXQW IDFWRU LV SURSRVHG WR
UHGXFHWKH ODUJHGHYLDWLRQVDV ZHOODVD VWHSIDXOWWR DFKLHYHWKH
PLQLPXPDV\PSWRWLFYDULDQFHFRQWURO%\DQDO\VLVWKHLQIOXHQFH
RI WKH IDXOW GLVFRXQW IDFWRU UHVHWWLQJ DQG IDXOW WROHUDQW 5)7
DSSURDFK LV SUHVHQWHG 6LPXODWLRQ VWXG\ VKRZHG WKDW WKH
SURSRVHGDSSURDFKLVHIIHFWLYH
,,1752'8&7,21
XULQJ WKH SDVW IHZ \HDUV PDQ\ UHVHDUFKHUV DQG
SUDFWLWLRQHUVKDYHORRNHGLQWRFRPELQLQJWKHDGYDQWDJHV
RIVWDWLVWLFDOSURFHVVFRQWURO63&DQGHQJLQHHULQJDXWRPDWLF
SURFHVVFRQWURO(3&$3&WKXVUXQ WRUXQ FRQWURODSSURDFK
KDV EHHQ SURSRVHG DQG ZLGHO\ DSSOLHG LQ VHPLFRQGXFWRU
PDQXIDFWXULQJ LQGXVWU\ ,W¶V D IRUP RI GLVFUHWH SURFHVV DQG
PDFKLQHFRQWUROLQZKLFKWKHSURGXFWUHFLSHZLWKUHVSHFWWRD
SDUWLFXODU PDFKLQH SURFHVV LV PRGLILHGH[ VLWX LH EHWZHHQ
PDFKLQH ³UXQV´ VR DV WR PLQLPL]H SURFHVV GULIW VKLIW DQG
YDULDELOLW\ >@ ([SRQHQWLDOO\ ZHLJKWHG PRYLQJ DYHUDJH
(:0$ DOJRULWKP LV RIWHQ DGRSWHG LQ UXQWRUXQ FRQWURO
%R[ DQG -HQNLQV >@ FDUULHG RXW SLRQHHULQJ ZRUN RQ WKH
(:0$ VWDWLVWLF DQG VKRZHG WKDW (:0$ SURYLGHV WKH
PLQLPXPPHDQVTXDUHHUURUIRUHFDVWIRUDQ,0$SURFHVV
$IWHU 6DFKV HW DO >@ LQWURGXFHG VLQJOH (:0$ LQWR
VHPLFRQGXFWRUPDQXIDFWXULQJLQGXVWU\DFRQVLGHUDEOHQXPEHU
RISDSHUVGLVFXVVHGWKHFKRLFHVRIWKHRSWLPDO(:0$ZHLJKW
,QJROIVRQDQG 6DFKV>@DQG 6PLWK>@ DQDO\]HGVWDELOLW\DQG
VHQVLWLYLW\ RI WKH SURFHVV RXWSXW IRU GLIIHUHQW FORVHGORRS
V\VWHPV-:DQJDQG43+H>@DQDO\]HGDQGFRPSDUHGWKH
0DQXVFULSWUHFHLYHG-XO\ WK7KLVZRUNZDV VXSSRUWHGLQSDUWE\
&KLQHVH1DWLRQDO1DWXUDO6FLHQFH)RXQGDWLRQ
% $L DQG < =KHQJ DUH ZLWK WKH &RQWURO 6FLHQFH DQG (QJLQHHULQJ
'HSDUWPHQW +XD]KRQJ 8QLYHUVLW\ RI 6FLHQFH DQG 7HFKQROXJ\ :XKDQ
&KLQDHPDLODLELQJELVKHQJ#IR[PDLOFRP
==KDQJLVZLWKWKH+XQDQ7HOHYLVLRQ6WDWLRQ&KDQJVKD&KLQD
HPDLO]]PDL#KRWPDLOFRP
< =KHQJ LV WKH FRUUHVSRQGLQJ DXWKRU SKRQH  ID[
HPDLO]\KLG\#PDLOKXVWHGXFQ
EHKDYLRUVRIVLQJOH(:0$FRQWUROOHUZLWKJDLQDQGLQWHUFHSW
XSGDWLQJ)RU WKH V\VWHP ZLWKRXW GULIW 67 7VHQJ HW DO >@
SURSRVHG YDULDEOH (:0$ FRQWUROOHU WR DFFHOHUDWH
FRQYHUJHQFHUDWHDWWKHEHJLQQLQJRIWKHSURFHVV
$OOWKHDIRUHPHQWLRQHGPRGHOVDUHEDVHGRQWKHDVVXPSWLRQ
WKDWWKHUHLVRQO\DVLQJOHSURGXFWPDQXIDFWXUHGRQRQHWRRO
ZKLFKLVIDUIURPUHDOLW\(GJDUHWDOUHYLHZHGWKHSUREOHPVRI
PL[HG SURGXFW UXQWRUXQ FRQWURO LQ D KLJKPL[ IDE >@ DQG
SURSRVHGMXVWLQWLPHDGDSWLYHGLVWXUEDQFHHVWLPDWLRQ-$'(
DOJRULWKP WR LVRODWH W\SLFDO W\SHV RI SURFHVV GLVWXUEDQFH
REVHUYHGLQVHPLFRQGXFWRULQGXVWU\>@2$9DQOLHWDO>@
SURSRVHG D PRGHO VHOHFWLRQ DSSURDFK WR LGHQWLI\ WKH FRQWH[W
YDULDEOHVWKDW FRQWULEXWH PRVWWRWKHSURFHVV 0' 0DHWDO
DQDO\]HG WRRO DQG SURGXFW HIIHFWV LQ D PL[HG SURGXFW DQG
SDUDOOHO WRRO HQYLURQPHQW >@ DQG SURSRVHG $129$
DSSURDFK WR GHDO ZLWK UXQ WR UXQ FRQWURO RI D KLJK PL[HG
RSHUDWLRQ >@ =KHQJ HW DO >@ VWXGLHG D PRGHO ZLWK WZR
SURGXFWV PDQXIDFWXUHG RQ WKH VDPH WRRO DQG SURSRVHG WZR
NLQGV RI FRQWURO PHWKRG ³WRROEDVHG´ DQG ³SURGXFWEDVHG´
DSSURDFKHV 7KH\ IRXQG WKDW WKH ³WRROEDVHG´ DSSURDFK LV
XQVWDEOHZKHQWKHSODQWLVQRQVWDWLRQDU\DQGWKHSODQWPRGHO
PLVPDWFK SDUDPHWHUV DUH GLIIHUHQW IRU GLIIHUHQW SURGXFWV
ZKLOH WKH ³SURGXFWEDVHG´ DSSURDFK LH WKUHDGHG(:0$
DSSURDFKLVVWDEOH:XHWDO>@JDYHWKHIXUWKHUH[SHULPHQWDO
VWXG\WRSURYH=KHQJ¶VUHVXOWV
)RUWKH³SURGXFWEDVHG´DSSURDFKLIWKHOHQJWKRIRQHWKUHDG
LVORQJLHORQJFDPSDLJQWLPHIRUWKHSURGXFWZKLFKLVEHLQJ
PDQXIDFWXUHGLQWKHWKUHDGWKHQRWKHUWKUHDGVZLOOKDYHDORQJ
EUHDN WLPH )RU WKH ORQJ WKUHDG DOWKRXJK WKUHDGHG(:0$
FRQWUROLVVWDEOHWKHRXWSXWRIRWKHUWKUHDGVZLOOEHIDUGHYLDWHG
IURP WDUJHW DW WKH EHJLQQLQJ UXQV RI HDFK F\FOH VHH WKH
H[DPSOHLQVHFWLRQ 7KHVHODUJHGHYLDWLRQV ZLOOEULQJKLJK
UHZRUNUDWHDQGORWVRIZDVWHVKRZHYHULWLVZHOONQRZQWKDW
PRGHUQ VHPLFRQGXFWRU PDQXIDFWXULQJ KDV HOHYDWHG FRVW LQ
SURGXFWLRQV LPSURYHPHQW LQ SURGXFWLRQ HIILFLHQF\ ZLOO
SRWHQWLDOO\EHYHU\EHQHILFLDOWRPDQXIDFWXUHUV,QWKLVSDSHU
WKHSUREOHPRIODUJHGHYLDWLRQVLQWKHILUVWIHZUXQVLVVWXGLHG
EDVHG RQ WKH WKUHDG(:0$ FRQWUROOHU &\FOHG UHVHWWLQJ
DOJRULWKPIRU GLVFRXQW IDFWRU LV SURSRVHGWRUHGXFHWKHODUJH
GHYLDWLRQV DW WKH EHJLQQLQJ UXQV RI HDFK F\FOH DV ZHOO DV WR
DFKLHYHWKHPLQLPXPDV\PSWRWLFYDULDQFHFRQWURO

$)DXOWWROHUDQW$OJRULWKPZLWK&\FOHG5HVHWWLQJ'LVFRXQW)DFWRULQ
6HPLFRQGXFWRU0DQXIDFWXULQJ,QGXVWU\
%$L<=KHQJDQG==KDQJ
'
2009 IEEE International Conference on Control and Automation
Christchurch, New Zealand, December 9-11, 2009 WeMT6.6
978-1-4244-4707-7/09/$25.00 ©2009 IEEE 483
,,352%/(0)2508/$7,21
,Q VHPLFRQGXFWRU PDQXIDFWXULQJ LQGXVWU\ PL[HG SURGXFWV
DUHXVXDOO\SURGXFHGZLWK WKH VDPH WRROV&RQVLGHUDVLPSOH
FDVHWKDW
S
SURGXFWVDUHPDQXIDFWXUHGRQDVLQJOHWRRO)LJ
7KHSURGXFWLRQVFKHGXOHFRQVLVWVRIF\FOHVRI LUXQVLQZKLFK
M
UXQV DUH XVHG WR SURGXFH SURGXFW  S
M
UXQV DUH XVHG WR
SURGXFHSURGXFWS
M
LVGHILQHGDVWKHFDPSDLJQOHQJWKIRU
SURGXFWDQG
LMLVFDOOHGWKHEUHDNOHQJWKRISURGXFW
)LJ$QH[DPSOHRISSURGXFWVDUHPDQXIDFWXUHGRQWKHVDPHWRRO
$VVXPLQJ WKDW WKH LQSXWRXWSXW UHODWLRQVKLS IRU WKH
SURGXFWV RQ WKH JLYHQ WRRO LV OLQHDU ZLWK GLIIHUHQW LQWHUFHSWV


S
αα α
DQG VORSHV 

S
ββ β
 DOO WKH SURGXFWV
VKDUHWKHVDPHWRROGLVWXUEDQFH LW Q
η
+






LW Q LW Q
LW Q LW Q
LW Q
SSLWQLWQS
XQM
XMQM
<
XMQL
αβ η
αβ η
αβ η
++
++
+
++
++
°++ +
°
=®
°
°++ +
¯

ZKHUH LW Q
X+LVWKHPDQLSXODWHGYDULDEOHDWWKHEHJLQQLQJRIWKH
()
WK
LW Q+UXQ WLV WKH QXPEHU RI F\FOH

LW Q
<Q M
+=LV
WKH RXWSXW RI SURGXFW  DW WKH HQG RI WKH
()
WK
LW Q+UXQ

LW Q S S
<QM M L
+=+ +LVWKHRXWSXWRISURGXFWSDW WKHHQG
RIWKH
()
WK
LW Q+UXQ ,QWKLVSDSHUZH DVVXPHWKDWWKHSURFHVV
GLVWXUEDQFH ^`
V
η
IROORZV DQ ,0$   VHULHV ZLWK
GHWHUPLQLVWLFOLQHDUGULIW
δ
LH
VV V V
ηεθεδ
−−
−= +

ZKHUH
V
ε
LV LQGHSHQGHQW LGHQWLFDOO\ GLVWULEXWHG UDQGRP
YDULDEOHDQG
a
V1
εσ

,Q³SURGXFWEDVHG´UXQWRUXQFRQWUROWKH(:0$ILOWHU
DFWLRQLV SHUIRUPHGZLWKUHVSHFWWR WKHODVWUXQRQ ZKLFKWKH
VDPHSURGXFWLVSURFHVVHG7KXVWKHRXWSXWRISURGXFWLQWKLV
FRQWURO DSSURDFK LV LUUHOHYDQW ZLWK ZKDW LV SURGXFHG E\ WKH
LMEUHDN UXQV )RU VXFK NLQG RI WKUHDGHG(:0$ FRQWURO
WKHSURFHVVZKLFKSURGXFHSURGXFWLVFRQVLGHUHGDVWKUHDG
$OVRWKHSURFHVVZKLFKSURGXFHVSURGXFWSLVFDOOHGWKUHDGS
:LWKRXWORVVRI JHQHUDOLW\IROORZLQJZHZLOOIRFXVHG RQ WKH
EHKDYLRURISURGXFW
*LYHQWKHSURFHVVSUHGLFWHGPRGHORISURGXFW
LW Q LW Q
<DEX
++
=+

ZKHUH Q=

M

D
DQG
EDUH PRGHO RIIVHW DQG JDLQ
SDUDPHWHUHVWLPDWHGIRUWKHV\VWHPUHVSHFWLYHO\
%\XVLQJ WKHWKUHDG(:0$ ILOWHUWKHSURFHVVRIIVHW LV
HVWLPDWHGWREH






LW Q LW LW L W M
LW Q LW Q LW Q
<EX Q DQGW
D<EX DQDQGW
<EX D Q M
λ
λλ
λλ
+++ +
++ +
−==
°
=−+ =
®
°−+ =
¯

WKHGHDGEHDWFRQWUROOHULV







LW M
LW Q
LW Q
7D QDQGW
E
7D
XQDQGW
E
7D QM
E
−+
+
+−
==
°
°
°
==
®
°
°=
°
¯

ZKHUH
7LVWKHGHVLUHG WDUJHW RI SURGXFW  DQG >@
λ
LV
FDOOHGGLVFRXQWIDFWRU
0100 200 300 400 500 600 700 800 900
−10
−5
0
5
10
15
Run
(a)
Output of product1 at cycle 0,1,2,3,4
λ
1
= 0.95
0100 200 300 400 500 600 700 800 900
0
2
4
6
8
10
12
Run
(b)
Output of product1 at cycle 0,1,2,3,4
λ
1
= 0.1
)LJ2XWSXWRISURGXFWLQF\FOH
6HW WKH WUXH SDUDPHWHU RI SURGXFW  WR EH
α
=
β
=

θ
= 

σ
= 
δ
 M= DQG L=
$VVXPH WKH OHDVW VTXDUH HUURU /6( RI 

α
β
LV

 DE =
DQG WKH WDUJHW YDOXH RI SURGXFW  LV 7=
)LJ VKRZV WKH RXWSXWV RI SURGXFW  E\ XVLQJ (:0$
FRQWUROOHU ,W¶V LQWHUHVWLQJ WR VHH WKDW ZKHQ GLVFRXQW IDFWRU

λ
=WKHVLPXODWHGSURFHVVRXWSXWZLOORVFLOODWHDVVKRZQ
LQ)LJDEXWLQ)LJEZKHQ 
λ
=WKHRXWSXWRIWKHV\VWHP
ZLWKRXWRVFLOODWRU\ EHKDYLRU)RUERWK FDVHVLW¶VFOHDU WKDWDW
WKH EHJLQQLQJ UXQV RI HDFK F\FOH HVSHFLDOO\ IRU W=
WKH SURFHVV RXWSXWV DUH IDU GHYLDWHG IURP WKH WDUJHW EXW
FRQVLGHUDEOHUXQVODWHU WKHRXWSXWVKLWWKHWDUJHWZLWKDVPDOO
ELDV,QRWKHUZRUGVWKHRXWSXWFRQYHUJHVWRWKHGHVLUHGWDUJHW
YHU\VORZO\ZKLFKZLOOUHVXOWLQDKLJKUHZRUNUDWH
484
,,,&<&/('5(6(77,1*$/*25,7+0)25',6&2817)$&725
6LQFH WKH GLVWXUEDQFH
{
}
V
η
VDWLVILHV  WKH IROORZLQJ
OHPPDVJLYHWKHRXWSXWRISURGXFW
/HPPD7KHRXWSXWRISURGXFWDWWKH WK
QUXQLQF\FOH 
LVVKRZQDVIROORZ

 

Q
QQN
QQNQN
N
<7 7
ϕ
α
ξ
εδ ϕ
γ
ϕε θε δ
−−
−−
=
=+++++ +
¦

ZLWK
E
β
ξ
= 
ϕξ
λ
=− 
 

Q
Q
γ
<
=®
¯

/HPPD 7KH RXWSXW RI SURGXFW  DW HDFK UXQ LQ WKH

WK
WWSURGXFWLRQF\FOHLVJLYHQE\










>   @

M
MW Q MW Q N
LW Q
N
M
MW Q N
MN MN
N
L
MW Q
NL M
NM
Q
QN
LWQN LWQN
N
<7 7
LM
7
ϕαξεδϕ ϕ ξλ
ϕϕεεδ
ϕδθεεε
ϕγϕεθε δ
+− −+ +
+
=
−+
−−
=
−+
+
=
+− +−−
=
=++++ +
+−+
+−+++
+− + +
¦
¦
¦
¦







^
 `
MM
W
PM
QNN
LW P M N LW P M N
PN N
L
LW P N LW P LW P M
NM
7
LM
χϕ ϕ ϕξλ ϕ ε ε δ
δθε ε ε
−−
−+
−−+− −−+−−
== =
−−+ + −−+
=
+++
+− + +− +
¦¦ ¦
¦

ZLWK
 

W
W
χ
<
=®
¯

$VPHQWLRQHGLQ6HFWLRQWKHSURFHVVRXWSXWLVIDUGHYLDWHG
IURPWKHGHVLUHGWDUJHW DWWKHEHJLQQLQJ UXQVRIF\FOH «
DIWHUDFRQVLGHUDEOHUXQVWKHRXWSXWVKLWWKHWDUJHWZLWKDVPDOO
ELDV,QWKLVVXEVHFWLRQ ZH ZLOO DQDO\VLV WKH PDLQ UHDVRQ IRU
WKLVSUREOHP
)RUDVWDEOHSURFHVVWKHPDWKHPDWLFH[SHFWDWLRQIRUSURGXFW
¶VRXWSXWDWWKH WK
QUXQLQF\FOH WWLVVKRZQDVIROORZ
()



Q
QQ
Q
(< 7 7
ϕ
ϕ
αξ ϕ δ
ϕ
−−
=+++

IRU W=DQG
()


Q
LW Q
(
<7 LM
δ
ϕ
δ
ϕ
+++−

IRU W
&RPELQLQJDQGZHKDYHWKHIROORZLQJOHPPD
/HPPD  7KH PDLQ ELDVHV RI SURGXFW ¶V RXWSXWV EHWZHHQ

WK
LW +DQG
()
 WK
LW M−+ UXQ
()
WK
LW Q+DQG
()
WK
LW Q+− UXQDUH


LW L W M
(
<(< LM
δ
+−+
−=



Q
LW Q LW Q
(
<(< L
M
ϕϕ
δ
++
−=

ZKHUH WDQG Q
7KHRUHWLFDO DQDO\VLV LQ VXEVHFWLRQ  LV FRQVLVWHQW ZLWK
VLPXODWLRQUHVXOWVLQ)LJLHLWZLOOWDNHDORWRIUXQVIRU
JHQHUDO (:0$ FRQWUROOHU ZLWK IL[HG GLVFRXQW IDFWRU WR
HOLPLQDWHWKRVHODUJHGHYLDWLRQVZKLFKZLOOEULQJKLJKUHZRUN
UDWH DQG SRRU HIILFLHQF\ ,Q RUGHU WR KDQGOH WKHVH SUREOHPV
F\FOHG UHVHWWLQJ &5 DOJRULWKP IRU GLVFRXQW IDFWRU LV
SURSRVHG
/HPPD,IWKHV\VWHPLVVWDEOHDV\PSWRWLFYDULDQFH$9DU
RISURGXFW¶VRXWSXWLQF\FOH WWLV
()

 
Q
$9DU <
ϕϕθ σ
ϕ
−−
=

IRU W=
()

LW Q
$9DU <
θθϕσ
ϕ
+
−+
=

IRU W
/HPPD  7R DFKLHYH PLQLPXP DV\PSWRWLF YDULDQFH
FRQWURORIWKUHDGGLVFRXQWIDFWRUVKRXOGEH
θ
λ
ξ
=
7KHRUHP&\FOHG5HVHWWLQJ$OJRULWKPIRU'LVFRXQW)DFWRU
)RUDVWDEOHSURFHVVGLVFRXQWIDFWRURIWKUHDGLQHDFKF\FOHDW
WKH WK
QUXQVKRXOGEH






 

JLW Q
LW Q
JLW Q
λλ
ξ
λ
λλ
ξξ
++ <
°
°
+=
®
°−+ <<
°
¯

ZKHUH 
J
LW Q+LVDGHFUHDVLQJIXQFWLRQRI QDQGVDWLVILHV
Q JLW Q+≥


JLW Q
λ
ξ
<+ +DQG



JLW Q
λ
ξξ
<− +<
6LQFHLQWKHILUVWIHZUXQVRIHDFKF\FOHWKHSURFHVVRXWSXWV
DUHIDURXWRIVSHFLILFDWLRQHVSHFLDOO\IRUF\FOH WWDWWKLV
WLPHPRUHHPSKDVLV VKRXOG EH JLYHQ WR WKH ODUJH GHYLDWLRQV
ZLWKWKHLQFUHDVHRI QWKHELDVEHFRPHVVPDOOHULQWKLVFDVH
ZHVKRXOGFRQVLGHUWKHPLQLPXPYDULDQFHFRQWURO

485
,9)$8/772/(5$7(&21752/
,Q VHPLFRQGXFWRU PDQXIDFWXULQJ LQGXVWU\ WKHUH DUH PDQ\
HYHQWVZKLFKFDQUHVXOWLQLPPHGLDWHIDXOW*HQHUDOO\ZKHQ
IDXOW KDSSHQV LV VHOGRP NQRZQ EXW WKH IDXOW ZLOO EULQJ
LQHYLWDEOH RVFLOODWLRQV DQG GHVWUXFWLRQV WR RXU SURFHVV
7KHUHIRUHWKHFRQWUROOHUVKRXOGOHDUQIURPWKHV\VWHPRXWSXWV
DQGWU\WRRYHUFRPHWKHIDXOW
,Q SUDFWLFDO SURGXFWLRQ WKH WRRO XVXDOO\ XQGHUJRHV D
PDLQWHQDQFHHYHQWVXFKHYHQWXVXDOO\FDXVHVDVKLIWZKLFKFDQ
EHVHHQDVDVWHSIDXOWWRWKHSURFHVV,QWKLVVHFWLRQGLVFRXQW
IDFWRUUHVHWWLQJDQGIDXOWWROHUDQW5)7DSSURDFKLVSURSRVHG
EDVHGRQWKHDQDO\VLVRIWKHSURFHVVZKLFKH[SHULHQFHVDVWHS
IDXOW
7KHVWHSIDXOWLVGHQRWHGDVLILWKDSSHQVDW WK
KUXQ


V
I
VK
I
V
K
=®<
¯

ZKHUH
I
LVWKHPDJQLWXGHRIWKHIDXOW
,I
V
)
GHQRWHVWKH VXPPDWLRQRIWRRO GLVWXUEDQFH
V
η
DQG
WKHVWHSIDXOW
V
I
DWWKH WK
VUXQLH
V
VV
)I
η
=+

WKHQZHKDYHIROORZLQJOHPPDV
/HPPD,QSUHVHQWRIVWHSIDXOWWKHPDWKHPDWLFH[SHFWDWLRQ
RI SURGXFW ¶V RXWSXW DW WKH WK
QUXQ LQ F\FOH LV
()



QQ
QQ QN Q
QNN
N
(< 7 7 I I I
ϕ
ϕϕαξ δγϕ ϕ
ϕ
−−
=
=− + + + + +
¦

3URRI&RPELQLQJDQGOHW W=WKHQ










QQ QQ
QQ
QN N
QN QN
NN
QQ
Q
N
QN QN QN QN
N
<< 7))
<7))
77I
II
ϕξ
λ
ϕϕξλϕ
ϕϕαξεδ
γϕ
εθε δ
−−
−−
−−
==
−−
−−
=
=++
=+ +
=− + + ++ +
+−++
¦¦
¦
7DNLQJPDWKHPDWLFH[SHFWDWLRQRI Q
<LQDERYHHTXDWLRQZH
JHW/HPPDƑ
/HPPD,QSUHVHQWRIVWHSIDXOWWKHPDWKHPDWLFH[SHFWDWLRQ
RISURGXFW¶VRXWSXW DWWKH WK
QUXQLQF\FOH WWFDQEH
GHVFULEHGDV
()







MN
M
QQ
LW Q L W N L W N
N
Q
QQN
LW L W M LW N LW N
N
(< 7 L M I I
II II
δϕδϕϕ
ϕ
ϕγϕ
−+
−−
+−++
=
−−
+−+ ++
=
≈+ + +
+−+
¦
¦

3URRI&RPELQHDQGWKHQ






LM LM
M
MM
N
M
NMN L M
N
<< 7))
7) 7 ) ) ) )
ϕξ
λ
ϕα ξ ϕ ϕ
++
+
−−−+
=
=+ +
=++++ +
¦
ZKHQ W=DQG Q=




LQ LQ LQ LQ
Q
QQ N
LLQNLQN
N
<< 7))
7< ))
ϕξ
λ
ϕϕ ϕ
++ ++
−−
+ +− +−−
=
=++
=− + +
¦
ZKHQ W=DQGQ












^
`

LW M
LW L W M LW
M
W
MW PM N
LLWPLWPM
PN
M
N
LW P M N LW P M N
N
MW M MW M
<< 7))
<7))
))
77)
ϕξλ
ϕϕϕξλ
ϕ
ϕϕϕαξ ϕ
−+
+−+ +
+−++
==
+
−−+− −−+−−
=
=++
=+ +
+−
=−++++
¦¦
¦








  

^
`
W
LM
MM
W
MW PM
NN
MN MN
NPN
M
N
LW P LW P M LW P M N LW P M N
N
))
)) 7
)) ) )
ϕϕ ϕϕξλ
ϕ
+
−−
−+
−−
===
+
+ −−+ −−+− −−+−−
=
+−+
+− +
¦¦¦
¦
ZKHQ WDQG Q=




LW Q LW Q LW Q LW Q
Q
QQ N
LW LW Q N LW Q N
N
<< 7))
<7))
ϕξ
λ
ϕϕγϕ
++ ++
−−
+++
=
=++
=++
¦
ZKHQ WDQG Q
&RPELQLQJDERYHHTXDWLRQVZHJHW










^
Q
QQ N
LW Q LW LW Q N LW Q N
N
Q
MW Q QN
LLWQNLWQN
N
M
W
PM
QNN
LW P M N LW P M N
PN N
<< 7 ))
<7))
7) )
ϕϕγϕ
ϕϕγϕ
χϕ ϕ ϕ ξλ ϕ
−−
+ + +− +−−
=
−+
+++
=
−+
−−+− −−+−
==
=++
=++
++
¦
¦
¦¦






 



`



M
LW P L W P M
MW Q M W Q MW Q Q
M
MW Q MW Q N
L M MN MN
N
QW
PM
NQ
LW Q N LW Q N
NP
))
7) 7
)) ) )
))
ϕαξ ϕ ϕ ϕ
ϕϕϕ
γϕ χϕ ϕ
=
−+ +
+− +− +−
−+ −+
+−
=
+− +−−
==
+−
=+++ +
+−+
+−+
¦
¦
¦


  
^
`
M
N
N
M
N
LW P M N LW P M N LW P LW P M
N
7
)) ))
ϕξλ
ϕ
=
+
−−+− −−+− + −−+
=
+−+
¦¦
¦
/HPPDFDQEHREWDLQHGE\WDNLQJPDWKHPDWLFH[SHFWDWLRQRI
486
LW Q
<+Ƒ
)URP HTXDWLRQV  DQG  ZH NQRZ WKDW LI WKH IDXOW
KDSSHQV DW WKH WK
KUXQ WKH LQIOXHQFH RI WKH IDXOW RQ SURFHVV
RXWSXWDWWKDWUXQLV
I
LHWKHVWHSIDXOWOHDGVWKHSURFHVV
I
DZD\ IURP WDUJHW DW WKH WK
KUXQ %\ PHDQV RI PDWKHPDWLFDO
LQGXFWLRQZH FDQHDVLO\FRQFOXGH WKDWLIWKH IDXOWKDSSHQVDW
WKH WK
KUXQ WKHQ LW ZLOO OHDG WKH SURFHVV
V
I
ϕ
DZD\ IURP
WDUJHW DW WKH
()
WK
KV+UXQ ,Q RUGHU WR RYHUFRPH WKH IDXOW
7KHRUHPLVSURSRVHG
7KHRUHP  'LVFRXQW IDFWRU 5HVHWWLQJ DQG )DXOW 7ROHUDQW
$SSURDFK,IWKHVWHSIDXOWKDSSHQVDWWKH WK
KUXQLQ F\FOH W
WKHQ WKH GLVFRXQW IDFWRU DW
()
WK
KV+UXQ LQ 7KHRUHP  LH
LW K V
λ
++ VKRXOGEHUHVHWWHGDV LW V
λ
+VRDVWRUHGXFH
WKHLQIOXHQFHRIVXFKIDXOW
3URRI %\ DQDO\VLV RI  DQG  ZH FRQFOXGH WKDW WKH
LQIOXHQFHRIWKHIDXOWLVSURSRUWLRQDOWR
ϕ
VRDQ\WLPHLIWKH
IDXOWKDSSHQV VPDOO
ϕ
LVH[SHFWHG WRUHGXFHWKHGHYLDWLRQV
FDXVHGE\VWHSIDXOW7KHSUREOHPRIKRZWRFKRRVHVPDOO
ϕ
LV WUDQVIRUPHG LQWR DQRWKHU TXHVWLRQ WKDW LV KRZ WR FKRRVH
GLVFRXQW IDFWRU LW Q
λ
+ 6LQFH KRZ WR FKRRVH LW Q
λ
+
KDVGLVFXVVHGLQ7KHRUHPLW¶VRPLWWHGKHUHƑ
96,08/$7,21678'<
,Q WKLV VHFWLRQ VRPH VLPXODWLRQ H[DPSOHV DUH SURYLGHG WR
LOOXVWUDWH IHZ DUJXPHQWV SUHVHQWHG LQ SUHYLRXV VHFWLRQV
:LWKRXW ORVV RI JHQHUDOLW\ ZH IRFXV RQ WZR SURGXFWV DUH
PDQXIDFWXUHG RQ WKH VDPH WRRO ZLWK F\FOH RI  UXQV LH
L=DQGWKHFDPSDLJQOHQJWKRISURGXFWLV$VVXPH
WKDW RWKHU SDUDPHWHUV DUH 

σ
= 
δ
= WDUJHW YDOXHV RI
SURGXFWDQGDUH 
 77 =
:KHQ RVFLOODWLRQ KDSSHQV WKH VLPXODWHG RXWSXW RI
SURGXFWDQGLQF\FOHZLWKIL[HGDQG&5GLVFRXQWIDFWRU
LV VKRZQ LQ )LJ ,Q WKLV VLPXODWLRQ 
θ
=− 
ξ
= DQG

ξ
= )URP /HPPD  ZH NQRZ WKDW

λ
=DQG

λ
= ERWK GLVFRXQW IDFWRUV ZLOO OHDG WKH V\VWHP VWDEOH
EXWZLWKRVFLOODWRU\EHKDYLRU:HFKRRVHIL[HGGLVFRXQWIDFWRU

λ
= DQG &5 GLVFRXQW IDFWRU  Q
LW Q
λ
+=
IRUWKUHDG)L[HGDQG&5GLVFRXQWIDFWRUVIRUWKUHDGDUH

λ
=DQG
 QM
LW Q
λ
+= UHVSHFWLYHO\ )URP
)LJLW¶VFOHDU WKDWWKHRXWSXWRIERWKWKUHDGVRVFLOODWHDWWKH
EHJLQQLQJUXQVRIHDFKF\FOHEXWWKHWKUHDGVZLWK&5GLVFRXQW
IDFWRUV SHUIRUP EHWWHU WKDQ WKH WKUHDGV ZLWK IL[HG GLVFRXQW
IDFWRUV'HWDLOFRPSDULVRQVRIWKHSHUIRUPDQFHVDUHVKRZQLQ
7DEOH
0100 200 300 400 500 600 700 800
−10
−5
0
5
10
15
20
25
Run
Output of product1 and product2 in cycle 0−3
product1 with fixed λ
1
product1 with CR λ
1
product2 with fixed λ
2
product2 with CR λ
2
)LJ2XWSXWRISURGXFWDQGLQF\FOHZLWKIL[HGDQG&5GLVFRXQW
IDFWRUIRUWKHSURFHVVZLWKRVFLOODWLRQ
7$%/(,
3(5)250$1&()25352'8&7$1',1&<&/( )257+(352&(66:,7+
26&,//$7,21
F\FOH    
IL[HG06(    
&506(    
IL[HG06(    
&506(    
IL[HGYDULDQFH    
&5YDULDQFH    
IL[HGYDULDQFH    
&5YDULDQFH    

,IDVWHSIDXOWKDSSHQVWKHSHUIRUPDQFHVRISURGXFWDQG
DUHLOOXVWUDWHGLQ)LJ)LJDVKRZVWKHSURFHVVGLVWXUEDQFH
ZKLFK IROORZV DQ ,0$ VHULHV ZLWK GHWHUPLQLVWLF OLQHDU
GULIW DQG D VWHS IDXOW ZKLFK KDSSHQV DW WKH WK UXQ ZLWK D
PDJQLWXGH RI  7KH RXWSXWV RI RYHUGDPSHG V\VWHP ZLWK
IL[HG DQG 5)7 GLVFRXQW IDFWRU DSSURDFKHV DUH JLYHQ LQ
)LJE )LJF LV WKH VLPXODWHG RXWSXW RI XQGHUGDPSHG
V\VWHPZLWKIL[HGDQG5)7GLVFRXQWIDFWRU)RUERWKILJXUHVLW
LVREYLRXVWKDWZKHQWKHIDXOWKDSSHQVLWWDNHVDORWRIUXQVIRU
WKHRXWSXWFRQYHUJHWRWKHVWHDG\VWDWHLIGLVFRXQWIDFWRURIWKH
WKUHDGLVIL[HGZKLOHIRUWKHWKUHDGZLWK5)7GLVFRXQWIDFWRU
WKHRXWSXWFRQYHUJHWRWKHVWHDG\VWDWHPXFKIDVWHU
487
0100 200 300 400 500 600
0
10
20
30
40
50
60
70
80
90
Run
(a)
IMA(1,1) with drift and step fault
0100 200 300 400 500 600
−5
0
5
10
15
20
25
30
35
40
45
Run
(b)
Output of product1 and product2 in cycle 0−2
product1 with fixed λ
1
product1 with RFT λ
1
product2 with fixed λ
2
product2 with RFT λ
2
0100 200 300 400 500 600
−30
−20
−10
0
10
20
30
40
50
Run
(c)
Output of product1 and product2 in cycle 0−3
product1 with fixed λ
1
product1 with RFT λ
1
product2 with fixed λ
2
product2 with RFT λ
2
)LJ2XWSXWRISURGXFWDQGLQF\FOHZLWKIL[HGDQG5)7GLVFRXQW
IDFWRUZKHQDVWHSIDXOWKDSSHQVDWWKHWKUXQ
9,&21&/86,21
,QWKLVSDSHU ZH VWXGLHG WKH SURFHVV ZLWK ,0$
GLVWXUEDQFHGHWHUPLQLVWLFOLQHDUGULIWDQGVWHSIDXOW:LWK
WKUHDGHG(:0$FRQWUROOHUWKHSURFHVVRXWSXWVDWHDFKUXQ
LQ HDFK F\FOH DUH GHULYHG %DVHG RQ V\VWHP SHUIRUPDQFH
DQDO\VLV VWDELOLW\ FRQGLWLRQV DQG WKH FDXVHV RI ODUJH
GHYLDWLRQV LQ WKH ILUVW IHZ UXQV RI HDFK F\FOH DUH JLYHQ
&\FOHG UHVHWWLQJ &5 DOJRULWKP IRU GLVFRXQW IDFWRU LV
SURSRVHGWRGHDOZLWKWKHODUJHGHYLDWLRQVDWWKHEHJLQQLQJ
UXQV RI HDFK F\FOH DV ZHOO DV WR DFKLHYH WKH PLQLPXP
YDULDQFH FRQWURO 7KH VLPXODWLRQ VWXG\ VKRZV WKDW WKH
DSSURDFKHVDUHYHU\HIIHFWLYH
5()(5(1&(6
>@ - 0R\QH ( ' &DVWLOOR $ 0 +XUZLW] 5XQWR5XQ FRQWURO LQ
VHPLFRQGXFWRUPDQXIDFWXULQJ)ORULGD&5&3UHVV
>@ *( 3%R[ * 0-HQNLQV ³)XUWKHUFRQWULEXWLRQVWR DGDSWLYHTXDOLW\
FRQWUROVLPXOWDQHRXV HVWLPDWLRQRI G\QDPLFVQRQ]HUR FRVWV´ %XOOHWLQ
RIWKH,QWHUQDWLRQDO6WDWLVWLFDO,QVWLWXWHYROSS
>@(6DFKV$+X$,QJROIVVRQ³5XQE\UXQSURFHVVFRQWUROFRPELQLQJ
63&DQGIHHGEDFNFRQWURO´,(((7UDQV6HPLFRQG0DQXIYROQR
SS
>@ $ ,QJROIVVRQ ( 6DFKV ³6WDELOLW\ DQG VHQVLWLYLW\ RI DQ (:0$
FRQWUROOHU´-4XDOLW\7HFKQROYROQRSS
>@ 7 6PLWK ³1RYHO WHFKQLTXHV IRU WKH UXQ E\ UXQ SURFHVV FRQWURO RI
FKHPLFDOPHFKDQLFDO SROLVKLQJ´ 06 7KHVLV (OHFWULFDO (QJLQHHULQJ
DQG&RPSXWHU6FLHQFH'HSW0,7&DPEULGJH0$
>@-:DQJ43+H³(:0$UXQWRUXQFRQWUROOHUZLWK JDLQ XSGDWLQJ
VWDELOLW\ DQG VHQVLWLYLW\ DQDO\VLV´ LQ $P &RQWURO &RQI  SS

>@ 67 7VHQJ $% <HK << &KDQ ³$ VWXG\ RI YDULDEOH (:0$
FRQWUROOHU´,(((7UDQV6HPLFRQG0DQXIYROQRSS

>@7)(GJDU6.)LUWK&%RGH90DUWLQHQ]³0XOWLSURGXFW5XQWR5XQ
&RQWUROIRU+LJKPL[)DEV´LQ$(&$3&$VLDKVLQFKX7DLZDQ
>@6.)LUWK:-&DPSEHOO$7RSUDF7)(GJDU³-XVWLQ7LPHDGDSWLYH
GLVWXUEDQFH HVWLPDWLRQ IRU UXQWRUXQ FRQWURO RI VHPLFRQGXFWRU
SURFHVVHV´,(((7UDQV6HPLFRQG0DQXIYROQRSS

>@ 2$ 9DQOL 16 3DWHO 0 -DQDNLUDP (' &DVWLOOR ³0RGHO FRQWH[W
VHFWLRQIRUUXQWRUXQFRQWURO´,(((7UDQV6HPLFRQG0DQXIYRO
QRSS
>@0'0D&& &KDQJ'6+:RQJ66- ,GHQWLILFDWLRQRI WRRODQG
SURGXFW HIIHFWV LQ D PL[HG SURGXFW DQG SDUDOOHO WRRO HQYLURQPHQW -
3URFHVV&RQWURO
>@0'0D&&&KDQJ66-'6+:RQJ³0L[HGSURGXFWUXQWRUXQ
SURFHVV FRQWURO ± $Q $129$ PRGHO ZLWK $5,0$ GLVWXUEDQFH
DSSURDFK´-3URFHVV&RQWUROYRO
>@ < =KHQJ 4+ /LQ '6+ :RQJ 66 - . +XL ³6WDELOLW\ DQG
SHUIRUPDQFHDQDO\VLV RIPL[HGSURGXFW UXQWRUXQ FRQWURO´- 3URFHVV
&RQWUROYROSS
>@ 0) :X :. /LQ &/ +R '6+ :RQJ 66 - . +XL ³$
)HHG)RUZDUG)HHGEDFN 5XQWR5XQ &RQWURO RI D 0L[HG 3URGXFW
3URFHVV6LPXODWLRQ DQG([SHULPHQWDO6WXGLHV´,QGXVWULDO(QJ &KHP
5HVYROSS
488
... Distributed control architecture is also widely used in wireless sensor network where each sensor node only takes care of its own computation and communication with gateway [42]. Another example of distributed control design can be seen in high-mixed semiconductor manufacturing run-to-run process control where the same kinds of products are grouped together, and the control actions are made based on the output of the latest product of the same kind instead of the output of the previous product (which may be different) [43][44][45][46][47][48]. ...
Article
Full-text available
Time delay is a common phenomenon in robotic systems due to computational requirements and communication properties between or within high-level and low-level controllers as well as the physical constraints of the actuator and sensor. It is widely believed that delays are harmful for robotic systems in terms of stability and performance; however, we propose a different view that the time delay of the system may in some cases benefit system stability and performance. Therefore, in this paper, we discuss the influences of the displacement-feedback delay (single delay) and both displacement and velocity feedback delays (double delays) on robotic actuator systems by using the cluster treatment of characteristic roots (CTCR) methodology. Hence, we can ascertain the exact stability interval for single-delay systems and the rigorous stability region for double-delay systems. The influences of controller gains and the filtering frequency on the stability of the system are discussed. Based on the stability information coupled with the dominant root distribution, we propose one nonconventional rule which suggests increasing time delay to certain time windows to obtain the optimal system performance. The computation results are also verified on an actuator testbed.
Thesis
Full-text available
Semiconductor manufacturing industry has elevated cost in productions. Improvement of production efficiency is always an important goal for manufacturers. Because of the high capital costs associated with the process equipments, it is a common practice in today’s semiconductor manufacturing to have many different products and processes run on each processing tool, i.e., high-mix manufacturing; However, in the area of semiconductor manufacturing process control, most of researches have been based on the assumption that there is only a single product in the manufacturing line, i.e., single product process, which is far from reality. Therefore, this thesis investigates the control methods for mixed product production. The main work for high-mix manufacturing can be summarized as follows: 1. We study a model with two products manufactured on the same tool with the same production schedule, i.e., in each production cycle, the same kind of product will be manufactured for the same batches, and propose “product-based” double exponentially weighted moving average (“product-based” dEWMA) approaches into the system which undergoes the disturbance that follows the first order integrated moving average (IMA(1,1)) with deterministic drift, and find that if the break length of the product is large, even if the dEWMA controller is adopted, the output of the system is also deviated far from the specification. 2. In order to improve the tool efficiency, in the real manufacturing processes, many different kinds of products are usually manufactured on the same tool. Therefore we study a more complicated case which assumes that a number of different kinds of products are manufactured on the same tool with variable manufacturing cycles, and the campaign length and break length of each cycle are also variable. We find that for mixed product drifted process, if the break length of a product is large, then at the beginning runs of each cycle, the process output will far deviate from the target value. We propose cycle resetting algorithm for discount factor of EWMA (CR-EWMA), and cycle forecasting EWMA (CF-EWMA) to reduce those large deviations; we also propose a discount factor resetting fault tolerant (RFT) approach and fault tolerant cycle forecasting EWMA (FTCF-EWMA) algorithm to handle the step fault, which is cause by the maintenance of the tool, of the system. Because in the semiconductor manufacturing batch processes, each step is a complicated physiochemical batch process, so generally it is difficult to perform measurements on-line or carry out the measurement for each run, and those combined with the fact that many process tools are not designed for the addition of in situ sensor, resulted in measurement taken less frequently than every run, or at stochastic runs. Therefore there will be delays in the feedback of the system. The effect of the delay on the stability of the system is an important issue which needs to be understood. Unfortunately, in the field of semiconductor manufacturing run-to-run control, few works are available for this issue, and therefore in this thesis, we investigate the stability of the system with metrology delay. The main works for delay system are concluded as follows: 1. For the single product process, we propose two kinds of controllers, EWMA-I and EWMA-II controllers, and the stabilities of systems with both controllers which undergo different kinds of metrology delays are investigated. Necessary and sufficient conditions for the stochastic stability are established. 2. For the mixed product process, EWMA-I and EWMA-II controllers are also proposed, and we extend the theorems of single product process to analyze mixed product process. The stability conditions for the mixed product process are also obtained.
Article
Full-text available
In the semiconductor manufacturing batch processes, each step is a complicated physiochemical batch process; generally it is difficult to perform measurements online or carry out the measurement for each run, and hence there will be delays in the feedback of the system. The effect of the delay on the stability of the system is an important issue which needs to be understood. Based on the exponentially weighted moving average (EWMA) algorithm, we propose two kinds of controllers, EWMA-I and II controllers for single product process and mixed product process in semiconductor manufacturing in this paper. For the single product process, the stabilities of systems with both controllers which undergo different kinds of metrology delays are investigated. Necessary and sufficient conditions for the stochastic stability are established. Routh-Hurwitz criterion and Lyapunov's direct method are used to obtain the stability regions for the system with fixed metrology delay. By using Lyapunov's direct method, the stability region is established for the system with fixed sampling metrology and with stochastic metrology delay. We also extended the theorems of single product process to mixed product process. Based on the proposed theorems, some numerical examples are provided to illustrate the stability of the delay system.
Chapter
Run-to-run (R2R) control is a form of adaptive model-based process control that can be tailored to environments where the process is discrete, dynamic, and highly unobservable; this is characteristic of processes in the semiconductor manufacturing industry. It generally has, at its roots, a rather straightforward approach to adaptive model-based control such as a first-order linear plant model with moving average weighting applied to adapt the (zeroth-order) constant term in the model. Most of the complexity of R2R control science lies and will continue to lie in extensions to support practical application of R2R control in semiconductor manufacturing facilities of the future; these extensions include support for weighting and bounding of parameters, run-time modeling of a large number of disturbance types, and incorporating prediction information such as virtual metrology and yield prediction into the control solution.
Article
The performance of a process control algorithm based on the EWMA statistic is analyzed. A simple condition relating the EWMA weight and the estimated process gain is shown to ensure that the control strategy is stable for a first-order multiple-input, single-output process that may be drifting or wandering. A similar condition is shown to guarantee stability for a second-order, single-input process. An expression is derived for the output mean squared deviation from target as a function of the algorithm parameters, how quickly the process drifts or wanders, and the amount of noise. The results are illustrated in the context of an epitaxial growth process.
Article
A novel run-to-run control algorithm based on a dynamic analysis of variance (ANOVA) approach is proposed to deal with run-to-run (RtR) control of a high mixed operation, i.e., many different products are manufactured in many different tools. The conditions of different tools and products are identified based on the ANOVA analysis of the system output. A dynamic term in the form of an autoregressive integrated moving average (ARIMA) disturbance model is included in the process model to characterize the run-to-run disturbances such as drift, shift and/or some other unknown disturbances of different tools. It is shown from the study below that controller performance can be improved by introduction of the dynamic term, especially for products which are produced only occasionally. This makes it highly suitable for mixed product control system. An industrial example is also included to demonstrate superiority of this approach.
Article
Run-to-run (RtR) control is an important quality assurance method for batch-based manufacturing process. Usually, products of different grades are produced on a tool that will experience gradual drift between maintenance cycles. A feed-forward/feedback RtR control strategy that compensates this drift for all products manufactured on this tool was proposed. This and other RtR control schemes were analyzed and validated by simulation and experimentally using a bench scale reactor that produces silica particles with different diameters by a sol−gel process. A simple EWMA (exponentially weighted moving average) RtR control scheme based on products of the same grade was found to be stable but inefficient for infrequent products. A simple EWMA RtR control scheme that attributed disturbance entirely as the effect of tool drift was found to be unstable. The feed-forward/feedback RtR control proposed was able to maintain stable quality by effectively utilizing information about tool changes to adjust recipes of infrequent products.
Article
Run-to-run control has been widely used in batch manufacturing processes to reduce variations. However, in batch processes, many different products are fabricated on the same set of process tool with different recipes. Two intuitive ways of defining a control scheme for such a mixed production mode are (i) each run of different products is used to estimate a common tool disturbance parameter, i.e., a “tool-based” approach, (ii) only a single disturbance parameter that describe the combined effect of both tool and product is estimated by results of runs of a particular product on a specific tool, i.e., a “product-based” approach. In this study, a model two-product plant was developed to investigate the “tool-based” and “product-based” approaches. The closed-loop responses are derived analytically and control performances are evaluated. We found that a “tool-based” approach is unstable when the plant is non-stationary and the plant-model mismatches are different for different products. A “product-based” control is stable but its performance will be inferior to single product control when the drift is significant. While the controller for frequent products can be tuned in a similar manner as in single product control, a more active controller should be used for the infrequent products which experience a larger drift between runs. The results were substantiated for a larger system with multiple products, multiple plants and random production schedule.
Conference Paper
Exponentially weighted moving average (EWMA) controllers are the most commonly used run-to-run controllers in semiconductor manufacturing. Based on a linear model, an EWMA controller is usually implemented in a way that the process gain is kept as the off-line estimate and the intercept term is updated using an EWMA filter at each run. However, in practice, there are many applications that an EWMA controller is implemented in a way that the process gain is updated in a run-to-run manner while the intercept is kept as the off-line estimate. Although the stability and sensitivity of EWMA controllers with intercept updating has been well known, there is no analysis result on the stability and sensitivity of EWMA controllers with gain updating. In this paper, we analyze the behavior of an EWMA controller with gain updating and compare it to that of an EWMA controller with intercept updating. Both stationary and drifting processes are considered, the expression of the process output are derived and the output variances for stochastic processes are evaluated. In addition, simulation examples are given to illustrate the analysis results.
Article
In the semiconductor manufacturing industry, production resembles an automated assembly line in which many similar products with slightly different specifications are manufactured step-by-step, with each step being a complicated physiochemical batch process performed by a number of tools. This constitutes a high-mix production system for which effective run-to-run control (RtR) and fault detection control (FDC) can be carried out only if the states of different tools and different products can be estimated. However, since in each production run, a specific product is performed on a specific tool, absolute individual states of products and tools are not observable. In this work, a novel state estimation method based on analysis of variance (ANOVA) is developed to estimate the relative states of each product and tool to the grand average performance of this station in the fab. The method is formulated in the form of a recursive state estimation using the Kalman filter. The advantages of this method are demonstrated using simulations to show that the correct relative states can be estimated in production scenarios such as tool-shift, tool-drift, product ramp-up, tool/product-offline and preventive maintenance (PM). Furthermore, application of this state estimation method in RtR control scheme shows that substantial improvements in process capabilities can be gained, especially for products with small lot counts. The proposed algorithm is also evaluated by an industrial application.
Article
Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1996. Includes bibliographical references (p. 139-143). by Taber H. Smith. M.S.