Moldflow MPI

CD-ROM

S/N

The mold flow analysis of the injection

molding process using Taguchi method and

grey relational analysis

Abstract

This article makes use of quality engineering program of the Taguchi method and the

grey relational analysis to develop the procedure about the many distinct engineering factors

effect the quality of objective on the injection molding process. We also collocates the mold

flow analytical software of Moldflow MPI to process the mold flow analysis.

Simultaneously, it analyzes and proves the influence of the many distinct engineering factors

on the quality of objective each other, from many experiment result of each engineering

factor's combination, to establish the optimal process of plastics injection molding in order to

execute the effective computer simulation. We take the warping phenomenon of CD-ROM

disk pallet for an example, in accordance with the process of variation of signal to noise ratio

and the grey relation analysis for each engineering factor upon the distribution of shear

stress, acquire that the result for the influences of the objective quality will identically. The

arrangement of effect magnitude is the temperature of melt, filling time, filling pressure and

temperature of mold. From the degree of contribution or relation we obtain that the

noticeable variations are the temperature of melt and filling time. Those factors will increase

the quality of product. The filling pressure and temperature of mold, the unnoticeable

variations, will be the basis to reduce cost.

Key words : Taguchi method, grey relational analysis, signal to noise ratio, degree of relation,

injection of form.Ko-Ta Chiang : Associate Professor, Department of Mechanical Engineering, HIT

[1,2]

(CAE)

(CAE)

C-mold moldflow Moldex

Kamal and Keing [3]

Wu et al. [4]

N o n -

Newtonian fluid Behavior

Hieber and Shen [5]

Hele-

S h a w [ 6 ] N o n -

Newtonian fluid

Chiang et al. [7]

Hele-Shaw[6] Hetu et al. [8]

3D

Pandelidis and Zou [9]

Choi et al. [10]

Neural network

(CAE)

CAE CAE

(try-and-error)

( Ta g u c h i

Method) [11]

1923 R. A. Fisher [12]

(orthogonal array) (signal to

noise ratio, S/N) (analysis of

variance, ANOVA) (response

table) (response graph)

(Taguchi Method)

[13,14]

[15,16]

[17]

( G r e y

system)

[18-22]

[23]

Laing[24] Chang

et. al. [25]

I-deas Master Series 8

Moldflow MPI

CD-ROM

(orthogonal array)

(signal to noise ratio,

S/N) (analysis of variance,

ANOVA) (response table)

(response graph)

(the larger-the-

better ) (the smaller-the-

better ) (the nominal-the-

better )

(

)

S/N S/N

M.S.D.

(the mean square

deviation)

S/N

S / N

Q(X) R

Xi

X0(k) Xi(k) i 0

Xi(k) X0(k)

Xi X0

[18-22]

1.

2.

3.

4.

Xi(k) X0(k)

[18-22 ]

ξ

0 1 0.5 [18-

22]

m i n

max

ri (k)

0.5

ri (k)

Xi Xo

r ( Xo Xi )

CD-ROM

ABS PA-746

187 127 18(mm)

1.5mm 3D

(shear stress)

CD-ROM

1.

2.

3.

ABS

L9

Moldflow MPI

S/N

yi n

S/N

A3 B1 C1

D 1 A 1

B2 C2 D3 S/N

S/N

0.5

Xi

Xo

r ( Xo Xi )

Xo

Xi r (

Xo Xi )

1.2565

S/N

92.64%

4.31%

I-deas Master Series 8

Moldflow MPI

CD-ROM

S/N

[1] J. Bown, Injection moulding of plastic

components, McGraw-Hill,

Maidenhead, pp.88-146, 1979.

[2] R. J. Crawford, Plastic engineering, 2nd

edition, Pergamon press, Oxford,

1989.

[3] M. R. Kamal and S. Kenig," The

injection molding of thermoplastics",

Polym. Eng. Sci. 12, pp.294-302,

1972.

[4] P. C. Wu, C. F. Huang and C. G. Gogos,

"Simulation of mold filling process",

Polym. Eng. Sci. 30(10), pp.882-892,

1990.

[5] C. A. Hieber and S. F. Shen, " a finite-

element/finite-difference simulation of

the injection molding filling process",

J. Non-Newtonian Fluid Mech. 7,

pp.1-32, 1980.

[6] H. Schlichting, Boundary-layer theory,

McGraw-Hill, New York, 1968.

[7] H. H. Chiang, C. A. Hieber and K. K.

Wang, "A unified simulation of the

filling and post filling stages in

injection molding. Part I.

Formulation", Polym. Eng. Sci. 31(2),

pp.116-123, 1991.

[8] J. F. Hetu, D. M. Gao, A. Garcia-Rejon

and G. Salloum, "3D finite element

method for the simulation of the

filling stage in injection molding",

Polym. Eng. Sci. 38(2), pp.223-236,

1998.

[9] I. Pandelidis and Q. Zou, "Optimization

of injection molding design. Part I.

Gate location optimizatiom, Polym.

Eng. Sci. 30(10), pp.882-892, 1990.

[10] G. H. Choi, K. D. Lee and N. Chiang,

"Optimization of process parameters

of injection molding with neural

network application in a process

simulation environment", Ann. CIRP

43(1), pp.449-452, 1994.

[11] G. Taguchi, Introduction to quality

engineering, Asian Productivity

Organiztion, Tokyo, 1990.

[12] R. A. Fisher, Statistical methods for

research workers, Oliver and Boyd,

London, 1925.

[13] D. C. Koa, D. H. Kimb and B.M.

Kimc, "Application of artificial neural

network and Taguchi method to

preform design in metal forming

considering workability",

International Journal of Machine

Tools & Manufacture 39, pp.771-785,

1999.

[14] W. H. Yang and Y. S. Tarng, "Design

optimization of cutting parameters for

turning operations based on the

Taguchi method", Journal of Materials

Process Technology 84, pp.122-129,

1998.

[15] A. Mertol, "Application of the Taguchi

method on the robust design of

molded plastic ball grid array

packages", IEEE Trans. Components

Packaging and Manufacturing

Technol, Part B, pp.734-743, 1995.

[16] R. S. Chen, H. C. Lin and C. Kung,

"Optimal dimension of PQFP by using

Taguchi method", Composite

Structures 49, pp.1-8, 2000.

[17] K. M. Tsai and P. J. Wang, "Semi-

empirical model of surface finish on

electrical discharge machining",

International Journal of Machine

Tools & Manufacture 41, pp.1455-

1477, 2001.

[18]

1-34

1985

[19]

24-55

1985

[20]

202-210

1986

[21]Deng, J., "Introduction to Grey System

Theory", The Journal of Grey System

1, pp.1-23, 1989.

[22]

15-20 1997

[23]

-

49-55

1994

[24]Laing, R. H., "Application of grey

relation analysis to hydroelectric

generation scheduling", Electrical

power and energy systems 21, pp.357-

364, 1999.

[25]Chang, S. H., Hwang, J. R. and Doong,

J. L., "Optimization of the injection

molding process of short glass fiber

reinforced polycarbonate composites

using grey relation analysis", Journal

of materials processing technology 97,

pp.186-193, 2000.

ETF

QQQ EWT

AMEX QQQ iShare EWT ETF

ETF ETF

ETF

ETF

Price Discovery and Market Integration of ETF

–An Empirical Study on QQQ & iShare EWT

listed in AMEX.

Abstract

This article uses daily closing price data of QQQ and iShare EWT listed in American

Exchange AMEX . To examine the process of the price discovery and market integration

between ETF and stock index. From the result of Cointegration Model, it reveals that there

has a long-term relationship between ETF and stock index. Two markets contribute a co-

integration system.

From the result of Error Correction Model ECM , it represents that the stock index

lead to ETF more stronger. In the short run, can know by impulse response analysis and

variance decomposition analysis, the stock index lead to ETF. But the relationship will tend

to convergence after six periods days .

Key words: Exchange Traded Fund ETF Price discovery Market integration

Cointegration Error Correction Model ECM

Ching jun Hsu : Associate Professor, Department of Financial Management, Nan-Hua UniversityPo hsin Wu : Postgraduate student, Department of Financial Management, Nan-Hua University

20 72 8

ROC ROC Fund

75

82

92

5 44

2.3 1

92 169

10

2

ETF

ETF3 Exchange Traded Fund

1993

AMEX

ETF S&P 500

SPDR S&P 500 Depositary Receipts

ETF

NAV

1021.43 1 70%

1. 2. 103. Exchange Traded Fund

AMEX ETF

AMEX

2 1996

Barclays Global

Investors BGI

Morgan Stanley Capital

International MSCI

ETF iShare MSCI 4

ETF

ETF

Sector/Industry Sector/Industry

ETF

4. WEBS World Equity Benchmark Shares

5

Pension fund

Poterba & Shoven 2002

ETF

6

90%

92 6 30 50

Taiwan Top-50 Tracker Fund

TTT 0050

ETF

5. ETF Large-capMid-cap Energy Utility Internet

6. James M. Poterba and John B. Shoven "Exchange Traded Funds: A New Investment Option for TaxableInvestors" NBER Working Paper No. 8781, February 2002

AMEX

ETF Nasdaq 100 QQQ

MSCI iShare EWT

ETF

Price discovery

Fama 1970

informationally

integrated Werner & Kleidon 1996

Market integration

ETF

ETF

ETF

ETF

ETF

1997

ETF

Chu Hsieh & Tse 1999 VECM

S&P 500

SPDR7

SPDR

Hasbrouck 2002

VECM

SPDR S&P 500

Intraday price formation

MDY8 S&P Midcap 400

S&P 500

ETF

S&P Midcap 400

ETF

ETF

ETF

Lai & Lai 1991 Johansen

Ghosh (1993)

S&P 500

S&P 500 Wahab

& Lashgari (1993) S&P

500 FTSE-100

(simultaneous)

Brockman & Tse

1995 Johansen

Booth So & Tse

1999

DAX

DAX

Kim

Szakmary & Schwarz 1999 VAR

S&P 500 MMI NYSE

composite

S&P 500 MMI

Min & Najand

1999

7. SPDR Standard & Poor's Depository Receipts 1993 S&P 500 ETFSPDR ETF 350

6. MDY S&P Midcap 400 ETF 2000 22 400

KOSPI 200

30

ETF9

Depository Receipt

2002 VAR

ADR

Impulse response analysis

Variance decomposition

C o m m o n

stochastic trend

ETF

Engle & Granger 1987

N o n -

stationary

Order

Johansen &

Juse l ius 1990

Trace test trace

Maximum eigenvalue statistic Max

ETF

Cointegration vector

Error correction

9. SPDR S&P 500 S&P 500 Depository Receipts

Restriction

ETF

Long-term error

correction

Short-term dynamic

Common stochastic trend

Engle

Granger 1987

Xt Yt

Zt-1=Xt-1- Yt-1

Error correction term 1 2

m n t 1 2

Xt

Xt YT

Yt ai bj t

Xt

ci di t

Yt

Yt

ETF

ETF

Variance of

forecast error

E T F

MSCI

0.991833 Nasdaq 100

0.845722

ETF

E T F

TTT

Nasdaq 100

QQQ MSCI

iShare EWT

AMEX ETF

ETF

QQQ

Yahoo 1999 3 10

2003 5 30 1103

iShare EWT

Yahoo 2000 6 25

2003 5 30 766

MSCI Taiwan Index iShare EWT

iShare

EWT MSCI Taiwan Index

3 Nasdaq 100

QQQ MSCI EWT

Nasdaq 100 ETF

QQQ Jarque-Bera

MSCI

ETF EWT

Jarque-Bera

ADF

-2.178228 -2.072228 -

2.586517 -2.687552 1%

ECM

Engle & Granger

1987

I(1)

I(0)

Cointegration vector

ETF

4 ETF

iShare EWT

R2 D-W

Spurious regression

Jarque-Bera

Intercept

Linear trend Augmented Dickey-

F u l l e r A D F

Harris McInish Shoesmith & Wood

1995 i =6

AIC

ADF

1%

Stationary

I(1)

Johansen

ETF

5 trace 99%

QQQ Nasdaq 100

r=0

1 r 1

Nasdaq 100 QQQ

trace Johansen

Max Johansen

r

n-r

Osterwald and Lenum

EWT MSCI Taiwan

Index trace 99%

r=0

1

r 1

EWT

1992

r 0

0 r 1

6 7

ECM

ETF

Common stochastic trend

Impulse response

a n a l y s i s Va r i a n c e

decomposition

3 Nasdaq 100

QQQ

Nasdaq 100 QQQ

MSCI EWT

MSCI

EWT

Variance of forecasting

error

i n n o v a t i o n E T F

innovation

8 Nasdaq 100

9 MSCI

99%

EWT

0.4% EWT

94.40734%

62% MSCI

5.59266%

37%

95%

QQQ

0~5% QQQ

18%

Nasdaq 100 81%

ETF

ETF QQQ EWT

ETF

Nasdaq 100

MSCI

Intuition

ETF

ETF

ETF

ETF

ETF

ETF

C h u

Hsieh Tse 1999 Joel Hasbrouck

2002 S&P 500

SPDR

ETF SPDR S&P

500

Nasdaq 100 MSCI

ETF QQQ EWT

QQQ EWT

ETF

ETF

Fleming Ostdiek &

Whaley 1996 Booth So & Tse

1999 Kim Szakmary & Schwarz

1999 Roope & Zurbruegg 2002

ETF

0.355%

2% 3%

ETF

ETF

2001 "

"

90 6

2001 "ETF( )

"

90 7

(2002)"ETF

"

91 9

2002 "

"

Booth, G. G., R. W. So, and Tse 1999 ,

"Price Discovery in the German

Equity Index Derivatives Markets"

The Journal of Futures Markets, 19,

619~643

Brockman, P. and Y. Tse 1995

"Information Shares in Canadian

Agricultural Cash and Futures

Markets." Applied Economics

Letters, 2, 335~338

Chu Q. C., G. W-L. Hsieh and Y. Tse

1999 , "Price Discovery on the

S&P 500 Index Markets: An

Analysis of Spot Index Index

Future and SPDRs" International

Review of Financial-Analysis, 8,

21~34.

Engel, R. E. and C. W. J. Granger

1987 , "Cointegration and Error-

Correction: Representation,

Estimation, and Testing"

Econometrica, 55, pp: 251~276

Fleming, J., B. Ostdiek and R. E. Whaley

1996 "Trading Costs and the

Relative Rates of Price Discovery in

Stock, Futures, and Option Markets"

The Journal of Future Markets. Vol.

16, No. 4, pp: 353~387.

Gastineau, Gary L. 2001 "Exchange

Traded Funds: An Introduction." The

Journal of Portfolio Management,

Spring 2001, Vol. 27, Number 3, pp:

88~96.

Ghosh, A. (1993), "Cointegration and Error

Correction Models: Intertemporal

Causality between Index and Futures

Prices." The Journal of Futures

Markets, 13, No.2, pp: 193~198.

Harris, F. H. deB, T. H. McInish, G. L.

Shoesmith, and R. A. Wood.

1995 "Cointegration, Error

Correction, and Price Discovery on

Informationally Linked Security

Markets." Journal of Financial and

Quantitative Analysis, 30,

pp563~579

Joel Hasbrouck 2002 , "Intraday Price

Formation in US Equity Index

Markets" New York University

Working Paper, Oct. 2002

Johansen, S. and K. Juselius 1990

"Maximum Likelihood Estimation

and Inference on Cointegration with

Application to the Demand for

Money." Oxford Bulletin of

Economics and Statistics, 52, pp:

169~209.

Kim, M., A. C. Szakmary, and T.V.

Schwarz 1999 , "Trading Costs

and Price Discovery across Stock

Index Futures, and Cash Markets"

The Journal of Futures Markets, 19,

475~489

Lai, K. S. and M. Lai 1991 "A

Cointegration Test for Market

Efficiency" The Journal of Futures

Markets, 11, 567~575

MacKinnon, J. G., 1991. Critical Values for

Cointegration Tests, New York:

Oxford University Press

Min, J. H. and M. Najand 1999 "A

Further Investigation of the Lead-

Lag Relationship between the Spot

Market and Stock Index Futures:

Early Evidence from Korea" The

Journal of Futures Markets, 19,

217~232.

Osterwald and Lenum, M. 1992 , "A

Note with Quantiles of the

Asymptotic Distribution of the

Maximum Likelihood Cointegration

Rank test Statistics", Oxford Bulletin

of Economics and Statistics 54, pp:

461~472.

Poterba James M. and John B. Shoven

2002 "Exchange Traded Funds: A

New Investment Option for Taxable

Investors" NBER Working Paper,

No. 8781, February 2002

Roope, M. & R. Zurbruegg 2002 "The

Intra-day Price Discovery Process

Between the Singapore Exchange

and Taiwan Futures Exchange." The

Journal of Futures Markets, 22,

No.3, pp: 219~240.

Werner, I. M., A. W. Kleidon 1996

"U.K. and U.S. Trading of British

Cross-Listed Stocks: An Intraday

Analysis of Market Integration."

Review of Financial Studies, 9, pp:

619~664

White 1980 , "Heteroskedasticity-

Consistent Covariance Matrix and a

Direct Test for Heteroskedasticity,"

Econometrica, Vol. 48, pp: 817~838

Cubes 2003 1 31

MSCI iShare 2003 2 7

http://www.ici.org Investment

Company Institute.

http://www.msci.com Morgan Stanley

Capital International

http://www.ishare.com iShare

http://www.amex.com

http://www.polaris.com.tw

9 Spearman

(β)

Mutual Funds Classification Schemes and

Performance Persistence

Abstract

The performance persistence is a very important factor for investors to invest mutual

funds. We have found a classification scheme that does not exhibit the problem of

performance reversals. Our sample includes 120 funds from 1998 to 2002. We show that the

relation between performance persistence and the standard deviation. Extremely,

performance reversals do not display in the bond funds. After we add a variable representing

the degree of a fund momentum strategy to the factor analysis, the previous performance

reversals are largely removed. This result shows that dynamic investment strategies should

also be included when determining funds.

Key words: Mutual fund, Classification, Performance, Persistence, Reversal, Momentum

Ching-Jun Hsu : Professor, Institute of Financial Management, Nan-Hua UniversityChih-Chien Chiang : student, Institute of Financial Management, Nan-Hua University

2002 12

500

Spearman

Sharpe(1966) Jensen(1968)

Carlson(1970) Williamson(1972)

Grinblatt Titman(1992) 5

Goetzmann Ibbotson(1994)

( ) Brown

Goetzmann(1995)

Malkiel 1995 Brown

Goetzmann 1995

Kahn Rudd(1995)

( 85 )

Spearman

Spearman

Sharpe(1966) Jensen(1968)

Carlson(1970) Williamson(1972)

Sharpe(1966)

1944 1963 34

Sharpe

Spearman

Carlson(1970) 57

Sharpe

Treynor

Williamson(1972)

1961 1970 180

( 82 ) Spearman

77 80 12

Sharpe

M.C.V

Sharpe

(reverse)

( 84 )

83 4 84 4

Spearman

Grinblat t

Titman(1992) 1974 1984

Goetzmann

Ibbotson(1994) 1976 1987

Jensen

t

Goetzmann Ibbotson

Jensen

Brown

Goetzmann 1995

Jensen

1980~1981 1987~1988

Malkiel 1995

70

80

86

Kahn Rudd(1995)

(Fixed-Income)

( 85 )

( 86 )

Sharpe

Spearman

(TEJ)

1998 1 2003 1

55

4 4 10

8

4

4 31

Sharpe

Spearman

Spearman

(

1 ) NAVp,t

Sharpe

(Reward to Variability

Ratio)

(TEJ)

Sharpe

Spearman

Spearman

Spearman

d=Xr-Yr

d 0 X

Y rs=1 X

Y rs=-1 -1≤rs≥1

t

n-2 t

Spearman

(

)

Carhart(1997)

PR1YR

SMB HML

SMB HML

b

β b>1

>1 b<1

β<1

rit

Spearman

Markowitz

Sharpe

Sharpe

(τ)

Sharpe

E-view

Jarque-Bera

85

0.95

35 0.95

( )

( 85 ) (

86 )

95%

90%( 12/120 )

Sharpe

1998 1

Spearman

66.67% ( 6/9 ) Sharpe

66.67 % ( 6/9 )

1998 ~1998

1999 ~1999 2001

~2001

4

Spearman

Sharpe

Spearman

88.89%( 8/9 )

Spearman

(rs)

Sharpe

( )

(

)

Carhart(1997)

(β)

2000

9209.48 4555.91

2000

(β)

12 Beta

Beta

Beta

Beta

b it>1 b it>1.2

9

bit<1.2

13

(β)

1. Brown, Stephen J., and William N.

Goetzmann (1995),"Performance

persistence," Journal of Finance 50,

679-698.

2. Carlson, R. S. (1970), "aggregate

performance of mutual fund." Journal

of Financial and Quantitative Analysis

5, 1-31.

3. Carhart, Mark M., (1997) On

persistence in mutual fund

performance, Journal of Finance 52,

57-82

4. Grinblatt Mark, Sheridan Titman

(1992),"The persistence of mutual

fund performance," Journal of Finance

47, 1977-1984.

5. Goetzmann, William and R. G. Ibbotson

(1994), "On winners repeat?"Journal

of Portfolio Management 20, 9-18.

6. Jensen, Michael (1968), "The

performance of mutual funds in the

period 1945-1964," Journal of Finance

23, 389-416.

7. Kahn, Ronald N., and Andrew Rudd

(1995), "Does historical performance

predict future performance?"

Financial Analysis Journal 51, 43-52.

8. Malkiel, B. G. (1995), "Return from

investing in equity mutual funds 1971

to 1991," Journal of Finance 50, 549-

572.

9. Sharpe, William F. (1966), "Mutual fund

performance," Journal of Business,

119-138.

10. Williamson, J. P. (1972), "measurement

and forecasting of mutual fund

performance: Choosing an investment

strategy," Financial Analysis Journal

28, 78-84.

11. (1997)

12. (1997)

13. (1996)

14. (1996)

15. (1993)

16. (1995)

A Study of The Function Model on Enterprise

Information Portal

Abstract

This main objective of the study was to assess definitions, concepts, and main

components of enterprise information portal architecture. With factor analysis, we found

extract that business image, news, human resources, products, consult service, trade message

help to address portal functions requirements. The results showed that business image and

human resources are different, and different kinds of corporate portals.

Key words: Enterprise information portal, function model

Shu-Hui Chuang : Instructor, Department of Industrial Management, HITHao-Hang Lan : Student, Department of Industrial Management, HITYu-Jr Lin : Student, Department of Industrial Management, HIT

1.

Portal

Site

...

ERP

SCM EIP

CRM ...

1.

2.

3 .

2.

2.1

(Chan and Chung, 2002)

(Eckerson, 2002; Chan and

Chung, 2002)

EIP

(Murray,1999)

(White,1999)

(White,1999)

(Monczka and Morgan, 2000; Carbon,

2000) Rezayat (2000)

(Shilakes and Tylman ,1998)

(Shilakes and Tylman,1998)

(Murray,1999;White,1999;Shilakes and

Tylman,1998)

...

(Murray,1999; Viador,1999)

e

...

( M u r r a y , 1 9 9 9 )

Firestone(1999)

2.2

(CRM)

(ERP)

Word Power Point...

( 2001)

(2001)

(1)

(2)

White(1999)

3.

5 1

3

...

4.

3

s t r a t i f i e d

random sampling

104

...

75

300

1 6

90 9 26

90 10

10

75 72 75

74 296 4

98.667

4.1

40

34%

31%

15% 13%

5% 2%

4.2

SPSS

Kaiser

1

0.4 6

Cronbach α 0.7

( 2 )

6 28

3

1.

.. . (

2 0 0 0 ) α

0.8679 α 0.8 0.9

2.

α 0.8282 α 0.8

3.

...

α

0.8078 α 0.8

4.

0.7096

0.8 0.7

5.

α 0.7852

0.8 0.7

6.

(

2000)

(

1999)

0.7628

0.8 0.7

4

P 0.00

P 0.865

1.

P 0.003<0.01

...

2.

P 0<0.01

5. 1.

2.

3.

...

1.

2.

3.

4.

[1] (2000)

[2] (2000)

[3] (2001) e

Internet Pioneer

[4] (2001)

[http://www.upromo.com.tw]

[5] Eckerson, W. (2002), Business Portals,

Drivers, Definitions, and Rules,

Lokaliseret 28 august 2002.

[6] Murray, G. (1999), The portal is the

desktop,[http://archives.groupcomputi

ng.com//index.cfm?fuseaction=viewar

ticle&ContentID=66].

[7] Firestone, J. M. (1999), "Defining the

Enterprise Information Portal", at

http://www.dkms.com/EIPDEF.html

[8] Shilakes,C.C. and Tylman,J. (1998),

Enterprise information portals,

[http://www.sagemaker.com/home.asp

?id=500&file=Company/WhitePapers/

lynch.htm].

[9] Viador, C. (1999), Realizing the vision

of information at your fingertips,

[http://www.viador.com/pdfs/EIP_whi

te_paper-1_99.pdf].

[10] White, C., (1999a), Decision

threshold, Intelligent Enterprise,

December, 16, 1999, p35-40.

[11] White, C. (1999b), Enterprise

information portal requirements,

[http://www.decisionprocessing.com/p

apers/eip2.doc].

[12] White, C. (1999c), The enterprise

information portal marketplace.

Decision processing Brief,

[http://www.decision.com/papers/eip1.

doc].

[13] Chan M., F. and Chung W. C. (2002),

A Framework to develop an enterprise

information portal for contract

manufacturing, International Journal

of Production Economics, Vol. 75, pp.

113-126.

[14] Monczka, R. M. and Morgan J. P.

(2000), Outsouring : key to many

competitive battles, Purchasing, Vol.

129(3), pp. 85-91.

[15] Carbon, J. (2000), What buyer look for

in contract manufacturers, Purchasing,

Vol. 129(4), pp. 33-38.

[16] Rezayat, M. (2000, The enterprise-

Web portal for life-cycle support,

Computer-Aided Design, pp. 85-96.

2

Discussion in Knowledge Worker and

Knowledge Analysis Based example on the

Machine Tool Industry

Abstract

The purpose of this research is to discover what knowledge skills a knowledge worker

should have by strategic implementation methods to generalize eight knowledge capabilities

in a knowledge worker, which are shipment knowledge, machinery maintenance knowledge,

electronic maintenance knowledge, computer application knowledge, documentation

management knowledge, assistance training knowledge, material management knowledge

and equipment operation management. In order to facilitate enterprises to set up learning

guidelines to solve issues in lack of knowledge workers and reduction of knowledge workers'

training period. For non senior knowledge workers perform knowledge work after taking

knowledge skills from other senior knowledge workers. Finally, this research is based on

preliminary data and secondary data to come up with a knowledge management

implementation module. And with current information the research will offer theoretic

suggestions and further directions for continuing studies.

Key words: knowledge worker, knowledge management, machine tool industry.

Chang Ting-Chang : Instructor, Department of Industrial Management, HITLin Yu-Xiang : Student, Department of Industrial Engineering & Mangement, HITFan Mei-Chen : Student, Department of Industrial Management, HIT

1.

Davis and Botkin 1994

Drucker 1992

Davenport et al. 1996

- -

Quinn 1992

response time

2.

2.1

Nonaka et al. 1991

tacit

explicit

Davenport et al.

1996

Drucker

1992

Ziddle,1998

Dove 1998

MBA

Thurow,1997 Nelson and Winter

1982

Lei 1997

Schon, 1987

Nonaka and Takeuchi, 1995;

Sveiby, 1997

2.2

Jauch and Cruck 1989

Rubenstein-Montano et al.

2001

Davenport 1997

1.

1

2

3

4

5

2.

3.

4.

5.

6.

7.

8.

9.

10.

Drucker 1999

1900

2000

Drucker

1

2

3

4

5

6

3.

11

Polkinghorne 1993

4.

4.1

NC

NC

NC

CNC

CNC

NC

NC CNC

4.2

CNC

M C

PIM

CNC

1996 NRS

MAP/3000

1999 ERP

Baan

4.3

1

4.4

Know-

How

(1) Know-How

(2)

(3)

(4)

(1)

(2)

(3)

(4)

(5)

(6)

4.5

Gary 2002

Bierly and

Chakrabarti 1996

Quintas et al. 1997;

Zack, 1999

Miles et al. 1998

" " " "

(1) Know-How

(2)

(3)

(4)

(5)

(6)

(7)

FAQs Frequently Asked

Questions

(8)

4.6

Drucker 1999

1

2

3

2

(1) a

b

c

d

(2) a

b

c

(3) a

b

c

d

(4) a

b

c

d

(5) a

b

c

d

2

5. 3

Demarest 1997

4

4

80%

2

15% Bierly and Chakrabarti 1996

Davenport & Prusak 1999

1990

Drucker 1999

Lebas 1999

6.

[1] Bierly, P. and Chakrabarti, A., "Generic

Knowledge Strategies in the US

Pharmaceutical Industry," Strategic

Management Journal, 1996,pp.123-135.

[2] Davenport, T. H., "Ten Principles of

Knowledge Management and Four Case

Studies," Knowledge and Process

Management, pp.187-208,1997.

[3] Davenport, T. H., Jarvenpaa, S. L. and

Beers, M. C., "Improving Knowledge

Work Process," Sloan Management

Review, Summer 1996, pp. 53-65

[4] Dove, R., "The Knowledge Worker,"

Automotive Manufacturing and

Production, 1998, pp. 26-28.

[5] Demarest, marc, "Understanding

Knowledge Management," Long Ronge

Planning, Vol.30, 1997, pp.374-384.

[6] Drucker, Peter F., "Knowledge-Worker

Productivity: The Biggest Challenge",

California Management Review, Winter

1999.

[7] Drucker, Peter F., Managing for the

Future, The 1990s and Beyond, Trumana

Talley Books, 1992.

[8] Drucker, Peter F., "The Future That Has

Already Happened," The Futurist,

November 1998.

[9] Fleck, J., "Informal Information Flow

and the Nature of Expertise in Financial

Services," International Journal of

Technology Management, 1996, pp. 104-

128.

[10] Gary, Hilson, "KM is a Strategy, not a

Technology," Cmmunications &

information management, 2002.

[11] Howells, J., "Tacit Knowledge

Innovation and Technology Transfer,"

Technology Analysis & Strategic

Management, 1996, pp.91-105.

[12] Johannessen, J. A., Olaisen, J. and

Olsen, B., "Information Management in

Negotiations: The Conditions Under

which it Could be Expected that the

Negotiation Partners Substitute a

Competitive Definition of the Situation

for a Cooperative one," International

Journal of Information Management,

1997, pp. 153-168.

[13] Nonaka, I., and Takeuchi, H., The

Knowledge-Creating Company, Oxford:

Oxford University Press, 1995.

[14] Nonaka, Ikujiro, " A Dynamic Theory

of Organizational Knowledge Creation,"

Organization Science, Vol. 5, 1994 , pp.

14-37.

[15] Miles, G., Miles, R. E., Perrone, V.

and Edvinsson L., "Some Conceptual

and Research Barriers to the Utilization

of Knowledge," California management

review, 1998, pp.281-288.

[16] Polanyi, M., The tacit dimension,

Ma:Gloucester, 1966.

[17] Quintas, p., Leferere, P. and Jones, G.,

" Knowledge Management: A Strategic

Agenda," Long Range Planning, 1997,

pp.385-391.

[18] Rolf, B., Profession, Tradition Och

Tyst Kunskap, Nya Doxa, Nora, Sverige,

1995.

[19] Sch0n, D., Educating the Reflective

Practioner, London: Jossey-Bass, 1987.

[20] Stewart, T. A., Intellectual capital: The

New Wealth of Organizations, London:

Doubleday, 1997.

[21] Sveiby, K. E., The New Organization

Wealth: Managing & Measuring

Knowledge-Based Assets, San

F r a n c i s c o : B e r r e t t - K o e h l e r

Publisher,1997.

[22] Thurow, L. C., The Future of

Capitalism, Nicholas: Breeley

publishing, 1997.

[23] Zack, M. H., "Developing a

Knowledge Strategy," California

Management review, 1998.

[24] Lebas, M. J., "Performance

Measurement and Performance

Management," International Journal of

Production Economics, 1999.

[25] Davis, S., and Botkin, J. The Monster

Under the Bed : How Business is

Mastering the Opportunity of

Knowledge for Profit, Simon and

Schuster, 1994.

[26] Quinn, J. B., Intelligent Enterprise,

New York: Free Press.

[27] Jauch, C., and Cruck, M., Strategy and

Business Policy, 3nded. McGraw-Hill,

1989.

[28] Rubenstein-Montano, B., Liebowitz,

J., and Buchwalter, J., "A System

Thinking Framework for Knowledge

Management," Decision Support

Systems, Vol. 31, 2001, pp. 5-16

A study on the cognitive preference

of safety colors in Twain

Abstract

The evidently high mortalily rate caused by occupational accidents and medicine

misuse in Taiwan shows that occupational safety and health regulations have not been

properly aware and obeyed. The study was undertaken to investigate the cognition and

preference of background color of hazard warning labels within the people in Taiwan. A

large amount of local subjects were polled with questionary related to perceptions of

different colors used in hazard warning . The results of the study show that when red color is

used people are better aware of hazard warning include danger, attention, mortal jeopardy,

warning, prohibit, caution, stop, radioactive danger, possible danger, etc. . The consensus

of people in Taiwan accords with our results. The results also suggest that black color should

be adapted to increase the color dimension of the hazard warning label.

Key words: elements of hazard warning, safety colors.

Hung-Wen Cheng : Instructor, Department of Industrial Management, HITChang-Yi Yang : Associate Professor, Department of Industrial Management, HIT

[1]

[2-3]

[4-5]

[6]

(CNS 9328 Z1024) [7]

(CNS 9328 Z1024)

[8-11]

(

)

(profile)

(symbolic icon)

(background color) (text)

(number)

(CNS 9328 Z1024)

(CNS 9328 Z1024)

[7]

.

.

.

.

.

.

.

.

.

ANSI Z53.1

ANSI Z53.1

a. RED - Danger, used for emergency

stops on machines and to identify

fire and protection equipment.

b. ORANGE - Identifies dangerous

parts of machines or energized

equipment.

c. YELLOW - Caution, identifies

physical hazards, see OSHA

1910.144.

d. GREEN - Used in safety and first aid

equipment.

e. WHITE AND BLACK - Used for

traffic and housekeeping.

f. PURPLE - Designates radiation

hazards.

ANSI Z53.1 ANSI Z535.1

( CNS 11295 ) 1

1. CNS 11295( ) C2. CNS( )

( CNS 9328 Z1024 )

[6]

[12-17]

a . b . c .

d. e. f.

g. h. i.

j. 10

10

[21]

1

10

1 10

10 .

. .

. .

. .

. .

.

[22]

2

cen t ra l

limit theorem

H0 0

0.01 0.25

= ( Z 2 )2

2 [23]

Xi Xi

V(X) Xi N

2149 0.01

0.25

3

16

(simple random

sampling)

[21]

2230

(2149 )

1

10

(Repeated

Measures)

10 10

10

H0 : 10

= 0.01

2 t

=0.01

10

A .

B. C. D.

E. F. G.

H. I . J .

10

10 (10

) H0

F P H0

2 F

10

10

5

(1)

(2)

(

)

(3)

5

( )

(1)

15

(2)

(3)

(4)

NSC90-

2218-E-164-002

1. 1981

2. (1997)

3. (2003) (

)

4. (1997)

5. (2003)

6. (1998)

7. CNS 9328

1987

8. (1992)

9. UN-Recommendations on the Transport

of Dangerous Goods (1992), 7th Revised

Edition, New York, Document NO.

ST/SG/AC 10/Rev.7.

10. UN-Recommendations on the Transport

of Dangerous Goods (1992), 10th

Revised Edition, New York.

11.

12. (1993)

13. (1995)

14. (1987) CNS

9328

15. (1987) CNS

9332

16. (1987) CNS

9331 1987

17. (1987) CNS

9330 1987

18. C.W. Emory and D. Cooper, Business

Research Methods, 4th ed. Richard

Irwin, 1991

19. Standards for Educational and

Psychological Tests and Manuals

Washington, D. C. American

Psychological Association, 1966

20. L. Lapin. Statistics for Modern

Business Decisions, 2nd ed. Harcourt

Brace Jovanovich, 1978

UL-94 DSC TGA

LOI

ammonium polyphosphate (APP) tripheny phosphate (TPP)

APP

Effects of Flame Retardants on

the Unsaturated Polyester

Abstract

A series of flame retardant containing unsaturated polyester have been developed. The

flame retardants include carbon black and phosphorous containing (ammonium

polyphosphate (APP) and tripheny phosphate (TPP)) materials. The effects of different flame

retardants on the unsaturated polyester were investigated by UL-94, DSC, TGA, LOI and

adiabatic bomb calorimeter. The results indicate that the flame retarded effect of APP type

flame retardant was better than the TPP ones as the phosphorous containing flame retardant

was used along. Moreover, the flame retarded effect was all improved as the phosphorous

containing flame retardant was mixed with carbon black. It indicates that synergistic effect is

existed as the carbon black used with phosphorous containing flame retardants, especially the

APP flame retardant.

Key words: unsaturated polyester, phosphorous containing flame retardant, carbon black,

synergistic effect.

Yeng-Fong Shih : Associate Professor, Department of Chemical Engineering, HITYih-Tyng Wang : Student, Department of Chemical Engineering, HIT

(1-4)

50

23.5 36.5

35

-(5-8)

(9-10)

1990 Dunlop Co.

(11)

1.

2.

(1)UL94

UL-94

124 12.7

10

1.After flame :

t1

2.After flame time :

t2

3.Afterglow : ,

,

4.Afterglow time :

t3

(2) (DSC)

TA Instruments

(V2.6D) ( 10 mg)

70 c.c. / min 10 /min

30 400

(3) (LOI)

0.5 Polymer

Laboratories

(FTA ) Nair(12)

17 L / min

(4)

(adiabatic bomb calorimeter)

(values of heat of

combustion HOC) IKA®-WERKE

(C4000)

Hubbard(13) ASTM 240D

0.5(14)

60

(benzoic acid)

-(26454 11) J g-1

3.04 MPa

(5)

TA Instruments

Modulated TGA 2950 thermal analyser

100 ml/min ( 10

mg) 10 /min

700

1.LOI UL-94

LOI APP(LOI 22.5)

TPP(LOI 20.5) RL00-

02(LOI 22)

LOI

UL-94

TPP

APP V-2 RL00-02

V-0 APP

TPP LOI

V-0

LOI UL-94

RL00-02 UL-94

RL00-02

UL-94

2.DSC

DSC

UP-A UP-R UP-RT UP-

RA UP-RA

APP RL00-02

UP-R UP-RT

TPP

RL00-02

(UP-RT)

UP-A UP-R UP-RT UP-RA

UP-RA

3.TGA

TGA DTG

400

( ) TPP

220 400

(

) APP

310 400

10% ( )

8% ( )

TPP

220 400

8% ( )

APP

310 400

20% ( )

DTG

TPP

APP

APP

APP

TPP

APP

20%

4.

APP

APP

DSC

DSC UP-A UP-R UP-

RT UP-RA UP-

RA

APP (UP-

A) UL-94 V-2

DSC LOI

22.5

TPP

UP-R UP-RT

UL-94 V-0

LOI 22 DSC

TPP RL00-02

UL-94 V-0 TPP

(UP-T) TPP

RL00-02 LOI

24 TPP RL00-02

UP-RA UL94 V-0

DSC

LOI

25.5 20%

APP

RL00-02

1. M. E. Hall, J. Zhang and A. R. Horrocks,

Fire Material, 1994, 18, 231

2. J.. Eichhorn, J. Appl. Polym. Sci., 1964,

8, 2497

3. R. N. Rothon and P. R. Hornsby, Polym.

Degrad. Stab., 1996, 54, 383

4. G. Matuschek, Thermochim. Acta, 1995,

263, 59

5. I. Kenji, T. Masakij and Y. Bunji, Jpn.

Pat., 09,208,731 , 1997

6. T. Masaru, Jpn. Pat., 09,316,257 , 1997

7. S. H. Chiu and W. K. Wang, J. Appl.

Polym. Sci., 1998, 67, 989

8. R. Xie and B. Qu, Polym. Degra. Stab.,

2001, 71, 375

9. J. Jang, H. Chung , M. Kim and H. Sung,

Polym. Test., 2000, 19, 269

10. S. R. Owen and J. F. Harper, Polym.

Degrad. Stab., 1999, 64, 449

11. "

" 85 9

67/68 20~35

12. C. P. R. Nair, G. Clouet and Y. Guilbert,

Polym. Degrad. Stab., 1989, 26, 305

13. W. Hubbard, D. Scott and G.

Waddington, Experimental

Thermochemistry. Ed. F. Rossini,

Interscience Publishers Inc., 1956, Vol.

1, Chapter 5.

14. A. Xu-wu, H. Jun and B. Zheng, J.

Chem. Thermodynamics, 1996, 28, 1115

A Study on Human Factors of Bench-Work

Stations

Abstract

The working-surface height of a bench-work table, which is one of the primary

equipments in bench work, has to be decided according to the concept of human-factor

engineering by taking the physical limits of human operators into account so that the

possibility of causing injury can be minimized and the safety requirements can be fulfilled.

This study bases on the analysis of anthropometry data measured from human subjects in the

practice of bench work and infers an appropriate value for the height of a bench-work table,

as well as the proper range of the working area. The result suggests a 93-cm height of a

bench table and a 120-cm distance between the bench vises. The study focuses on the

students in the department of mechanical engineering, and the subjects include the first three

grades of students in a senior vocational/industrial school, the first four grades of students in

a five-year institute of technology, and the first-grade students in a two-year institute of

technology.

Key words: Human factors engineering, Anthropometry, Bench work.

Teh-Tsang Tsai : Instructor, Department of Mechanical Engineering, HIT

1/4 6 [1]

1/12 [2]

1/7 [3]

2mx1.2m 150mm[1]

[4]

(Flexion)

[5]

[5]

(Human factors)

[6]

(Anthropometry) 16

19

[4]

(Sta t ic

anthropometry) (Dynamic

anthropometry)

1.

(1).

(2).

2.

(1).

(2).

(3).

(4).

(5).

3.

4.

(Bench work)

(Bench vise)

50~80mm 1[7,8]

7 5 1 0 0 1 2 5 1 5 0 m m

(CNS4037,CNS4038)

300mm

10mm

45

2(a)

30 75

2 ( b ) 30~40 [7,8]

16 205mm

10mm

10mm

3

45

20 ~30 25

4

30

90

40 [7,8]

5~10mm

300mm

50~60 [7,8]

( W )

1220mm(4 ) 1500mm 1820mm(6 )

2 0 0 0 m m 2 1 2 0 m m 2 1 5 0 m m

2350mm 2430(8 ) (D)

700mm 760mm 820mm 910mm(3 )

970mm 1100mm 1210mm(4 )

1220mm 1300mm (H)750mm

760mm 800mm 820mm 860mm

900mm 970mm 1040mm 1080mm

3 6 3 7 4 8

CNS

125mm

145mm 170mm

175mm 180mm 185mm 190mm

205mm

[9]

5

10mm

10mm

95%

10 ~15

6

(Grandjean,1988) [5,10,11,12,13,14]

5~10cm

10~15cm

15~20cm[5]

9 0 ~ 9 5 c m [ 11 , 1 2 ]

Sanders and McCormick

88~107cm[13]

[9]

7 Barnes

Squires

[12,13]

288.3mm 565.1mm[16]

[4,17,18]

[19]

[12]

(Christensen,1988) [20]

84

[5,19]

(Sanders, McCormick ,1987) [4]

(Chandra Pinnagoda ,1996) [21]

[22,23]

[24]

18 25

[15]

1.

2.

3.

16~19

3 12

170 172 174 168 180 181

175 180 168 180 175 170cm

( ) 4.7 95%

5%

1

(n) [25]

2001 10

22 88

(Martin system

anthropometer)

1 2

[23,26]

1. (Stature)

2. (Span)

3. (Elbow height)

4. (Foot breadth)

5. (Foot length)

6. (Shoulder breadth)

7.

[24]

16 19 95%

114 .0cm

3.5cm

= + =114.0

3.5 117.5cm

5~8cm [7,8]

=

( 5~8cm)

125mm

180mm

=117.5 (18 5~8)

94.5~91.5cm 93cm

(Space bubble)

0.67

35cm

(

3 ) = = 3 5

49=84cm

3

( 8

A)

= ( sin45 )

=11 32 (49 sin45 )=76.7cm

( 8

B)

=

tan30 tan30

=76.7 tan30 tan30 =25.5cm

= 8 4

25.5=109.5cm

5%

95%

10cm

120cm

[5,9,10,23]

= =(183

49)/2=67cm

10cm[23]

=

(67-10) 2=114cm

122cm 243cm(4

8 )

9

10~15cm (toe

space) [27]

1.

87 8 76

2.

83 21 133~147

3.

83 21 117~118

4.

8 9

7,26,50~52

5.

8 7

109,112,114,116,117,120,128

6.

85

2

7.

8 7

127~128,133~136,156~157,174~175

8. Labour Department for Industrial

Professional Education: Basic

proficiencies metal working-filing,

sawing, chiselling, sharing, scraping,

fitting. Labour Department for

Industrial Professional Education.

1958, p.02-02-12-2, 02-02-23-2, 02-03-

07-2, 02-03-32-3.

9.

65

129~130,134

10.

86

22~23,26,47

11.

86 1

12. Christopher D. Wickens, Sallie E.

Gordon and Yili Liu: An introduction

to human factors engineering.

Addison-Wesley Educational

Publishers Inc., New York, 1998, p.2,

pp.315-316.

13. Mark S. Sanders and Ernest J.

McCormick: Human factors in

engineeering and design. McGraw-Hill,

Inc., New York, 1993, p.418, p.432,

pp.435-437.

14.

89 2-

33~35

15.

1

16.

23 2

17.

1995 43~52

18. Dan Macleod: The ergonomics edge.

Van Nostrand Reinhold, New York,

1995.pp.34-36.

19. Alphonse Chapanis: Human factors in

systems engineering. John Wiley &

Sons, Inc., New York, 1996. pp. 11-16.

20. Robert W. Proctor and Trisha Van

Zandt: Human factors in simple and

complex system. Allyn and Bacon,

Boston, 1994, p.3,p.389

21.

1998 9

22.

1

23.

8 8 6 1 ~ 6 9 , 7 3

122,124

24.

91

92 61

25.

91

299

26.

2 0 0 0

62~65

27. K. H. E. Kroemer, H. B. Kroemer and

K. E. Kroemer-Elbert: Ergonomic.

Prentice-Hall, Inc., New Jersey, 1994,

p.47.

T-S

T-S

(LMI)

T-S

T-S Fuzzy Modeling and Control for Electric

Vehicle Propulsion Using Linear Matrix

Inequality

Abstract

This paper presents a fuzzy control design approach which can meet the speed tracking

requirement when electric vehicles are operated on various traffic conditions. A T-S fuzzy

model for approximating the state equation of an electric vehicle propulsion system with

high energy efficiency-based ac motor drive is first proposed, and a robust fuzzy control

based on the T-S fuzzy model is then considered. The robust stabilization for the EV(electric

vehicle) propulsion system is cast into a linear matrix inequality (LMI) problem via roust

performance analysis, and the LMI problem can be solved efficiently by using the convex

optimization techniques. Computer simulations are presented for illustrating the performance

of the suggested control strategy.

Key words: EV propulsion, T-S fuzzy model, robust fuzzy control, robust stabilization, linear

matrix inequality, convex optimization techniques.

Cheng-Da Hsieh : Instructor, Department of Electrical Engineering HIT.

1. IntroductionOne dominant issue of EV design is to

lengthen the running distance on one

battery charge. This implies the importance

of a high-efficiency motor drive. As

enhanced insulated gate bipolar transistors

(IGBT's) are used in the PWM inverter, the

loss of the inverter is negligible. Mutoh et

al. [1] proposes an energy saving strategy

for induction motors. The rotor and stator

copper losses and the core loss are

minimized, and the optimal ratio of

magnetizing current to the torque current is

derived.

The stability and robustness problem

of the EV propulsion control systems is

also an important topic. Variable structure

control using a proper switching law can

drive the system into the predetermined

sliding mode, and according to the sliding

mode, the system can approach to its

equilibrium. Thus, the sliding mode control

approach can offer many good properties,

such as insensitivity to parameters

variation, external disturbance rejection,

and fast dynamic response [2]. One more

interesting design method of stabilization

strategy for complex nonlinear systems can

be as follows: first build a T-S fuzzy model

for approximating the nonlinear plant, and

a fuzzy model-based controller is then

synthesized utilizing the concept of

"parallel distributed compensation". Local

linear feedback controls can be designed

systematically by a generalized Lyapunov

function and some linear matrix

inequalities, and the closed-loop fuzzy

control system composed of the fuzzy

model and the PDC controller is globally

asymptotically stable [3,11].

In this paper, based on the optimal

relationship of the magnetizing current and

the torque current found by the energy-

saving control principle [1], the dynamics

model and a T-S fuzzy model for EV

propulsion systems with 3-phase AC

induction motor (IM) are first constructed.

Then, the parallel distributed compensation

(PDC) approach [3] is adopted for

synthesizing a robust speed tracking control

for EV propulsion systems. Through the

robust performance analysis for disturbance

rejection the control problem is translated

into a linear matrix inequality problem. By

considering it as a generalized eigenvalue

minimization problem (GEMP), the

required common Lyapunov matrix and the

local feedback gain matrices can be

decided. The derived LMI-based control

law can guarantee the stability and

robustness of an EV propulsion system

when it is driven from standstill to stable

cruise under various uncertainties.

The paper is organized as follows:

Section 2 presents the dynamics model of

an energy-saving EV propulsion system. In

Section 3, a T-S fuzzy model for the EV

propulsion system is proposed, and a LMI-

based robust speed control design using the

derived T-S fuzzy model is suggested.

Some representative simulation results are

shown in Section 4. Finally, conclusions are

made in Section 5.

2. Modeling of an Electric VehiclePropulsion System

Consider the power train for an EV

shown in Fig.1, where an ac induction

motor is used for generating the driving

torque. The dynamics model for the load

part consisting of reduction gears and a

differential gear for driving the rear wheels

of an EV will be derived by the first

principles. The ac induction motor is

assumed with the energy saving driving

strategy proposed by [1], and the complete

mathematical model for the power train

will be constructed.

The angular displacements of the

motor, the rear shaft and the rotor of the

rear wheels are defined as θm, θ1 and θ2,

respectively, as shown in Fig. 1. Let the

transmission ratio of the reduction gear

be N0=θm/θ1, and the reduction ratio for the

differential gears be Nd=θ1/θ2. Then the

load torque (Tl) equation can be derived

below.

Refer to Fig. 2, the equivalent

propulsion force Fp generated by the EV

drive system and the equation of motion of

the vehicle can be expressed as :

where T2 is the driving torque on the rear

wheel shaft (as shown in Fig. 1), and r is

the wheel radius. The inertia resistance far

can be derived as:

where v is the velocity of the vehicle, Jw =

Jw1 + Jw2 ( Jw1 and Jw2 are respectively the

moments of inertia of the front and rear

wheel shaft systems) is the total moment of

inertia of the rotation parts, and m = mB

+m1 +m2 is the total mass, here mB , m1 , and

m2 are respectively the mass of the vehicle

body, the front and rear wheel shaft

systems.

The aerodynamic drag is given as:

where ζ is the air density, Cw is the

aerodynamic drag coefficient, A is the

vehicle frontal area, Vo is the head-wind

velocity, v+Vo is the velocity of the vehicle

relative to the head-wind, and sgn is the

sign function. The grade resistance is

where αs is the grade angle, αs is positive

for the upgrade case and negative for the

downgrade case. The rolling resistance frr

composed of the front and rear parts ( frr,1

andfrr,2 ) can be expressed as:

where K is the tire

rolling resistance coefficient [6,7], and is

the gravity constant. By Eq. (1), we can

obtain

Since (refer to Fig. 1)

we have

where

here p=d/dt ; are the stator's

input voltage components along the q- and

d- axes of the rotor flux frame,

respectively; are the stator's

flux linkage components along the q- and

d- axes of the rf frame, respectively; rs and

rr are the resistances of the stator and rotor,

respectively; ωrf is the rotating velocity of

the rf frame; ωr is the rotating velocity of

the rotor; are the rotor's current

components along the q- and d- axes of the

rf frame, respectively; are

the rotor's flux linkage components along

the q- and d- axes of the rf frame,

respectively; Te is the electromagnetic

torque of the motor; P is the number of

poles; Lm is the magnetization inductance;

Lr is the rotor's inductance; and Lls and Llr

are the leakage inductances of the stator

and rotor, respectively.

By the principle of field orientation

and letting the d- axis be entirely aligned

with the rotor flux, we have, ,and

the torque equation can be simplified as

In order to generate the motor torque

with maximum efficiency, the total loss Pl

in the drive system must be a minimum.

The total loss Pl generated in the drive

system of an electric vehicle can be

summarized as follows [1]:

where

here rm is the core loss resistance, PSTR is

the stray load loss, PMEC is the

mechanical loss, and PINV is the inverter

loss. Only the losses Pl1 and PINV can be

controlled in the ac drive design. The loss

PINV is very small in comparison with the

loss Pl1, as long as enhanced insulated gate

bipolar transistors (IGBTs) are used in the

PWM inverter. In this case, the efficiency

of the inverter is generally more than 95%.

Thus, the PINV loss can be neglected.

Since , by Eqs. (12) and (17),

we have

Usually, the response of is much

slower than those of

almost equals zero. Hence Pl can be

simplified as

Substituting (16) into (18), we have

where is the motor torque

constant.

Let α be the ratio of the magnetizing

current to the torque current, i.e.,

Then Pl1 can be expressed in terms of α :

The optimal ratio that makes the

loss P1 a minimum can be derived by

letting :

Substituting

into the motor torque equation, we have

The dynamics model for the EV propulsion

system shown in Fig. 1 can thus be derived

as:

where ωm is the angular velocity of the

motor rotor in rad/s, Jl is the moment of

inertia of the motor shaft including the

reduction gear in this side. By substituting

(8) into (30), we have the complete

dynamics model as follows:

3. Fuzzy Model-Based ControlDesign of Electric VehiclePropulsion System

The LMI techniques will be applied to

the stability analysis of an electric vehicle

propulsion control system. Defining the

state variables as:

can be rewritten as:

where

αs and Vo are considered as with

uncertainty. After some manipulations, (32)

can be rewritten as

where is the state

vector; is the input

variable;

with

and and pu

including all the other terms is considered

as uncertainty. Without loss of generality,

we consider the major operating case of

sng(v+Vo)=1 and sng v=1

Since B is only a constant matrix, the

derivation of a T-S fuzzy model and then

the control design and finding of the

solution of LMIs become much less

difficult than the approaches directly based

on the original complex model.

Assume for the

nonlinear terms in Eq. (33). Defining z=X2,

the maximal and minimal values of z can

be deduced and expressed as below:

Choose z as the antecedent variable of the

T-S fuzzy model, we can define two fuzzy

sets with membership functions shown in

Fig. 3 in the universe of discourse of z.

Then a T-S fuzzy model can be constructed

analytically as follows:

Model Rule i:

where

The overall equation of the T-S fuzzy

model can be inferred as

where

here Ci (z) is the grade of membership of z

in fuzzy set Ci. By the membership

function definitions shown in Fig. 3, we

have

For arbitrary trajectory tracking

control, first define the tracking error

vector as

where is the

desired trajectory vector. Notice that

are the desired angular

displacement and velocity trajectories of

the rotor about its rotating axis,

respectively. Differentiating Eq. (38), we

have

Substituting (39) into the model rules (34),

we can obtain the T-S fuzzy model for the

error dynamics as follows:

Error Rule i :

The output of the error T-S fuzzy model

can be defuzzified as:

In this study, the PDC(Parallel Distributed

Compensation )[3,11]with control rules

constructed based on the T-S fuzzy model

rules is adopted for the fuzzy control

design. Each control rule has a linear state

feedback part and a feedforward part to

compensate for the effect of gravity, that is,

Control Rule i :

where So the design

objective is to determine the local feedback

gains Ki in the consequent parts of the

control rules via LMI.

The output of the PDC controller can be

inferred as:

By substituting (43) into (41), and since B

is a constant matrix in the T-S fuzzy model

and , the error dynamics for

the whole closed-loop system can be

derived as

For the consideration of robustness

with respect to the disturbance Pu , the

following robust performance requirement

[4] for the tracking error is to be met:

where , and is a symmetric

positive definite matrix. Equation (45)

means that the effect of Pu on the error

must be attenuated below a prescribed level

ρ. To synthesize the fuzzy controller that

can reject the external disturbances of an

electric vehicle propulsion control system,

we can select a positive definite function as

follows:

The requirement (45) for a prescribed

ρ>0 can be shown to be equivalent to the

following condition:

By integrating (47) from 0 to tf with initial

condition e (0)=0,we have

Thus,

Equation (49) implies (45). Therefore

if (47) holds, the robust performance

requirement can be guaranteed under Pu .

The LMI constraints can be derived

from (47). First, rewrite (47) as

and substituting (44) into (50),we have

That is,

Therefore, if the following constraints are

satisfied:

then Equation(52) holds. Employing the

Schur complements for nonstrict

inequalities [5],(53) becomes

Conditions (54) can be solved by

considering it as a generalized eigenvalue

minimization problem (GEMP), that is, to

maximize α subject to the following

constraints:

Because the second inequalities in

(55) are not jointly convex in P and Ki , it

is difficult to find a common solution P and

Ki . Fortunately, the inequalities can be

transferred into matrix inequalities by

variable transformation.

Defining new variable X=P-1, and

multiplying the inequalities on the left and

right by X, we can obtain

Equation (56) can be rewritten as

where Mi ≡ KiX. That is, the PDC control

design problem can be transformed to the

problem of maximizing α subject to the

following linear matrix inequality

constraints:

If there exists a common X and Mi 's

satisfying the above LMI constraints, then

the common P and Ki can be obtained as

There exist methods in the literature for

solving the LMI problems, such as interior

point algorithm [5]. The MATLAB

software package has incorporated this

algorithm into the solver of LMI control

toolbox. In this study, the instruction gevp

is used to solve the above GEVP problem.

Based on the proper choice of suitable α, P

and the feedback gains Ki,i=1,2, can thus

be determined, and the design of the

control law (43) is accomplished.

4. Simulation ResultsIn this section computer simulations

are used to illustrate the performance of the

proposed T-S fuzzy model-based control

strategy for an electric vehicle which is

operated on various traffic conditions. The

nominal values of the parameters used in

the simulations are chosen as:

(1) Induction motor (60Kw, 2430rpm,

250N-m) and transmission with the

following parameters:

(2) Road and load with following

parameters:

where ωf =209.44 rad/sec is the desired

maximum angular speed in 10<t≤14sec,

and tf = 10 sec.

Two traffic conditions are considered.

Traffic condition (I) is selected as:

Using the usual LMI method, the

feedback gain matrices and the

common positive definite matrix P can be

obtained as follows:

.Simulation results

for the case with EV operated on traffic

condition (I) are shown in Fig. 5. From Fig.

5(a), we know that the EV speed response

can track the command trajectory. The

tracking error shown in Fig. 5(b) is within

and

The corresponding required motor control

torque is shown in Fig. 5(c). The control

torque is smaller

than the maximum torque of the induction

motor.

Simulation results for the case with

EV operated on traffic condition (II) are

shown in Fig. 6. From Fig. 6(a), we know

that the EV speed can also follow the

command trajectory. The tracking error

shown in Fig. 7(b) is within -0.0023 and

0.0016 rad/sec. The corresponding required

motor control torque is shown in Fig. 6(c).

The control torque

is smaller than the maximum torque of the

induction motor.

From the simulation results for this

case with EV operated on the more traffic

condition, we know that the speed tracking

error with the LMI method is small and

from Fig. 5(c), we know that control torque

with the LMI method is smooth. Thus,

when ρ is selected as small as possible, and

the common P and the feedback gains

are found with a bigger α, So

the LMI method has excellent capability to

reject disturbances and high robustness

with respect to uncertainty.

5. ConclusionsIn this paper, the dynamics model and

a T-S fuzzy model for EV propulsion

systems with a 3-phase ac induction motor

are constructed. A procedure for

systematically constructing a simple T-S

fuzzy model with very small number of

rules that can exactly represent the EV

propulsion systems with a 3-phase ac

induction motor is suggested, and a PDC

control design based on the T-S fuzzy

model is proposed. Because the number of

rules is very small, it is easy to find a

common Lyapunov matrix P, and no

relaxation methods are need. The feedback

gains Ki and P can be simultaneously

determined by considering the control

design problem as a GEMP problem via

LMI constraints. Proper Ki and P can be

obtained by choosing the results with

sufficiently high value of the Lyapunov

function decay-rate scaling factor α .

Simulation results are used to show that the

derived LMI-based control law can

guarantee the stability and robustness of an

EV propulsion system when it is driven on

various traffic conditions.

References[1] N. Mutoh, S. Kaneko, T. Miyazaki, R.

Masaki, and S. Obara, "A Torque

Controller Suitable for Electric

Vehicles," IEEE Trans. Ind. Electron.,

Vol. 44, No. 1, pp. 54-63, 1997.

[2] K. K. Shyu and H. J. Shieh, "A New

Switching Surface Sliding Mode Speed

Control for Induction Motor Drive

Systems," IEEE Trans. on Power

Electronics,Vol. 11, No. 4, pp. 660-666,

1996.

[3] H. O. Wang, K. Tanaka and, M. F.

Griffin "An Approach to Fuzzy Control

of Nonlinear Systems: Stability and

Design Issues," IEEE Trans. on Fuzzy

System, Vol. 4, No. 1, pp. 14-23, 1996.

[4] Tseng, C. S., Chen, B. S., and Uang, H.,

J., "Fuzzy tracking control design for

nonlinear dynamic system via T-S

fuzzy model," IEEE Trans. on Fuzzy

Systems, Vol. 9, No. 3, pp. 381-392,

1995.

[5] S. Boyd, L. El Ghaoui, E. Feron, and V.

Balakrishanan, Linear Matrix

Inequalities in System and Control

Theory, Philadelphia, PA: SIAM, 1994.

[6] M. Ehsani, K. M. Rahwan, and H. A.

Toliyat, "Propulsion System Design of

Electric and Hybrid Vehicles," IEEE

Trans. on Industrial Electronics, Vol.

44, No. 1, pp. 19-27, 1997.

[7] B. K. Powell, K. E. Bailey, and S. R.

Cikanek, "Dynamic Modeling and

Control of Hybrid Electric Vehicle

Powertrain Systems," IEEE Control

Systems Magazine, Vol. 18, No. 5, pp.

17-33, 1998.

[8] D. W. Novotny and T. A. Lipo, Vector

Control and Dynamics of AC Drives,

Oxford 1996.

[9] B. K. Bose, Modern Power Electronics

and AC Drives, Prentice Hall PTR,

Upper Saddle River, NJ, 2002.

[10] J. J. Craig, Introduction to Robotics,

Addison- Wesley, 1989.

[11] K. Tanaka and H. O. Wang, Fuzzy

Control Systems Design and Analysis,

Wiley-Interscience, 2001.

(Ritz)

Numerical Solution of Buckling and Vibration

in Laminates with Arbitrary Shape Cut Off

Regions

Abstract

The pure global buckling and vibration of four sides simply-supported as well as

clamped anisotropic laminates having an arbitrary shape cut off region that is symmetric with

respect to mid-plane have been studied by treating the remaining cut off regions as uniform

plates with reduced stiffness. The variation of stiffness of the plate is represented by Fourier

series. Computational solutions of the energy principle for the Ritz method in a plate having

arbitrary shape of typical cut off regions under biaxial compressive loads are obtained. Some

numerical results for the pure global buckling load prediction due to its reduced flexural

stiffness for the circular cut off regions and elliptical cut off regions are presented. We find

the normalized pure global buckling load ratio decreases as the cut off regions size increases,

and the nondimensional fundamental frequency value decreases as the cut off regions size

increases.

Key words: pure global buckling, cut off region, reduced stiffness, arbitrary shape.

C. C. Hong : Assistant Professor, Department of Information Management, HITH. W. Liao : Assistant Professor, Department of Information System, Ling Tung CollegeM. F. Hwang : Associate Professor, Liberal Art Center, Da Yeh UniversityK. C. Jane : Professor, Department of Applied Mathematics, National Chung Hsing University

1. IntroductionMany researchers have studied

various aspects of buckling load and free

vibration of plates with and without the

delaminations analytically and

experimentally [1-5]. There are three

possible buckling mode types: (a) local

buckling mode, (b) global buckling mode

and (c) coupled global and local buckling

modes that have been examined. Cut off

regions on the top and bottom surfaces of

laminated plate may be necessary for fitting

something on it. Cut off regions may

reduce bending stiffness of laminates,

which lower the compressive load carrying

capacity and natural frequencies. In 2000,

Jane and Hong [6] made a study about the

pure global buckling and vibration of

rectangular laminates with rectangular cut

off regions. An energy approach for the

Ritz procedure was discussed by Whitney

[7] is used to determine the pure global

buckling load and vibration of four edges

simply-supported as well as clamped

anisotropic rectangular laminates that

having arbitrary shape cut off regions.

The stiffness variation of the plate is

represented by Fourier series [5] and the

partition technique [8] can be utilized to

approach the arbitrary shape of cut off

regions into the rectangular subregions and

right triangular subregions. The Ritz

method using the reduced flexural

stiffnesses to represent the stiffness of the

arbitrary shape of cut off regions under

biaxial compression loads would be

studied. The purpose of this study is to

investigate the effect of arbitrary shape cut

off regions to the pure global buckling and

vibration of rectangular plates by energy

method. The typical anisotropic rectangular

laminates with middle-plane symmetric

arbitrary shape cut off region that is shown

in Figure 1. With coordinates X1(x1,y1),

X2(x3,y1), X3(x3,y4), X4(x2,y3) and X5(x1,y2)

and is used to demonstrate the study

procedures. We partition the arbitrary shape

of cut off region into the rectangular

subregions and right triangular subregions

as shown in Figure 2. Where ζ(k) is the x-

coordinate at middle point of 2c(k) ,2c(k) ,

which is the x-directional length of cut off

sub-region and η(l) is the y- coordinate at

middle point of 2d(l) ,2d(l) , which is the y-

directional length of cut off sub-region.

2. Formulation2.1 Governing equation

The strain energy of an elastic plate in

terms of Cartesian coordinates x, y system

is written in the following relationship [7]:

where σx and σy are the normal stresses, σxy

is the shear stress, εx and εy are the normal

strains, and εxy is the shear strain. This

strain energy generally contains two parts

of energy, there are strain energy due to

stretching and strain energy due to bending.

By substituting the plane stress constitutive

equations and the strain-displacements

relations into relationship equation (1), we

find that the strain energy for pure

transverse bending of anisotropic laminated

plate can be written in the following

equation:

where W is the transverse displacement,

are the flexural stiffnesses.

The potential energy of external

inplane loads due to a transverse deflection

is given as follows:

where are initial

external inplane force resultants applied to

the rectangular plane in a prebuckled state,

are the midplane strains

due to the transverse deflection. We

consider the initial axial loads acting on the

plane in the x- and y- directions, these

external loads are represented as:

For considering the large transverse

deflection in the buckled state, we have

nonlinear terms in the strains involving

transverse displacement W as follows:

By substituting equations (4) and (5)

into equation (3), we arrive at the following

potential energy equation:

The kinetic energy of an elastic plane

in terms of a Cartesian coordinate x, y, z

system is written in the following

relationship:

where is the density of the k-th layer in

laminated plate, u,v and W are the

displacement components in the x, y and z

directions respectively, t is the time. For

considering the tangential displacement u,v

are linear functions of the z coordinate and

neglecting the rotatory inertia terms, after

integrating with respect to z, we have the

following expression:

where ρ is density of the laminated plate, h

is thickness of the rectangular plate, and

u0,v0 are the displacement components of

the mid-plane. Conventionally, we consider

the vibration under the following

displacement forms:

where ω is a natural frequency of vibration.

By putting equation (9) into equation (8),

the kinetic energy can be rewritten as:

2.2 Energy principle

The energy principle for the Ritz

procedure can be stated as [7]:

where Π is Lagrangian functional, U is the

bending strain energy, V is the potential

energy of inplane loads, and T is the kinetic

energy of the laminated plate. By

considering bending and linear inertia of

the plate, equation (11) becomes:

where a and b are dimensions of the

rectangular plate,

are the flexural stiffnesses and are

the bending-twisting coupling stiffnesses of

the laminate, Px and Py are applied biaxial

loads in x- and y- directions respectively.

2.3 Reduced flexural stiffness

When the arbitrary shape of cut off

region is occurred in the rectangular plate,

the overall effective flexural stiffness

would be smaller than the flexural stiffness

of perfect plate. We would like to partition

the arbitrary shape of cut off region into

sufficient numbers of rectangular shape of

subregions and right triangular shape of

subregions to get the approximate solution.

A double Fourier series form of reduced

flexural stiffnesses had been

used in the delaminated plate study by

Wang et al [5]. With {D} representing

original stiffnesses and

and the distribution function of

reduced flexural stiffnesses is written in the

following form:

where are the Fourier

coefficients, they are written in the

following forms for the arbitrary shape of

approximately cut off region that is shown

in Figure 2:

where is the

area of the cut off rectangular sub-

region of dimensions with

its center at as shown in Figure

2, KL is the total number of cut off

rectangular subregions, A=ab , is the

ratio of the amount of flexural stiffnesses,

same for all components, for the kl cut off

sub-region to the corresponding stiffnesses

of the no cut of laminate,ζ 0 =c 0 =0 ,ζ 2 =a,

c 2 =0, η 2 =b,d 2=0, K is the total number of

cut off rectangular sub-regions in the x-

direction. At each , there are L k cut off

rectangular subregions, L is the total

number of cut off rectangular sub-regions

in the y-direction. At each η (l) , there are K

l cut off rectangular subregions. is the

area of the cut off right triangular

subregion of dimensions

with at the middle point of

sidelong edge, TKL is the total number of

cut off right triangular subregions, TK is

the total number of cut off right triangular

sub-regions in the x-direction. At each

, there are TLk cut off right triangular

subregions, TL is the total number of cut

off right triangular subregions in the y-

direction. At each η (l), there are TK l cut off

right triangular subregions. For the closely

representing f(x, y) to actual stiffness, a

sufficient number of terms of a ij would be

used. Of course, only a 00 =1 exists if there

is no cut off region.

2.4 The Ritz method

The Ritz method provides a

convenient method for obtaining

approximate solutions for buckling and/or

vibration problems. For the present

problem of four sides simply supported,

and four sides clamped rectangular

anisotropic plates with arbitrary shape of

cut off region, the solution is assumed in

the following form:

and for the reduced flexural stiffnesses

in cut off plate, with the

process of minimization equation [7]:

Substituting equation (15) in conjunction

with equation (13) into equation (16), we

arrive at:

in which f (x,y) represents the distribution

function of bending stiffness and bending-

twisting coupling stiffness of the plate.

Although f (x,y) could be different for

different stiffness components, it is taken to

be uniform for all in the present study.

After integrating equation (17) with

properly assumed X m (x) and Y n (y), we

have the following system of equations.

in which contains

with specified as parameters. By

requiring the determinant of the coefficient

matrix in equation (18) to vanish, we have

the eigenvalue problem for the critical

under a given

2.5 Simply-supported rectangular plate

We now consider the simply

supported rectangular laminated plate

compressed by uniform inplane loads of

with specified The boundary

conditions for a four edges simply-

supported rectangular plate are written as

follows:

and the following characteristic functions

for are selected.

By substituting the equation (13) and (21)

into equation (17), and performing

integrations afterwards, we can obtain the

equations in series forms.

2.6 Clamped rectangular plate

We consider a four sides clamped

rectangular laminated plate compressed by

uniform inplane loads of with specified

The boundary conditions are:

and the following characteristic functions

for are selected.

By substituting the equation (13) and (24)

into equation (17), and performing

integrations afterwards, we also can obtain

the equations in series form.

3. Some numerical results anddiscussions

We assumed that there were no local

buckling occurred in the anisotropic

laminates under the whole processes of

axial compression in the numerical

simulations. The stiffness of a remaining

middle-plane symmetrical cut off region is

determined by the flexural stiffness

between every two adjacent cut off region,

through the thickness of the plate. For

example, if there is a middle-plane

symmetrical single cut off region through

the thickness as shown in Figure 3, we have

Firstly, we study the

convergence of pure global buckling

solution for a simply supported plate with

arbitrary shape of typical cut off region as

shown in Figure 1 with the coordinates

and

is shown in Figure 4a. And convergence of

pure global buckling solution for a clamped

plate with the same coordinates of arbitrary

shape of typical cut off region is shown in

Figure 4b. We find the total number of

terms resulting in a M=N=5 of [A] matrix is

established explicitly and used to

demonstrate the calculating procedure. The

typical stiffness for the perfect part is taken

to be

for all numerical

computations. Secondly, we study the

following cut off region cases:

3.1 Typical cut off region case

3.1.1 Simply supported plate

A plate with arbitrary shape of typical

cut off region (see Figure 1) with the

coordinates under biaxial loading condition

with is considered. The

following buckling load parameter is

introduced: The result

for the critical load of a plate without cut

off region by using the Ritz method is

found to be 3.0396, which should be the

maximum upper bound for all other

numerical results presented in this paper.

A square plate has arbitrary shape of

typical cut off region with y-coordinates

cut

off region position is occurred at =0.3.

Results on the pure global buckling load

normalized with respect to =3.0396

versus c1/b (c1=0.1b with x1=0.4b, x2=0.5b,

x3=0.6b; c1=0.2b with x1=0.3b, x2=0.5b,

x3=0.7b; c1=0.3b with x1=0.1b, x2=0.5b,

x3=0.9b) are shown in Figure 5. The result

show that normalized pure global buckling

load ratio decreases as the cut off

region size increases.

A square plate has arbitrary shape of

typical cut off region with the coordinates

Results on the pure global

buckling load normalized with respect to

versus are shown in

Figure 6. The results show that normalized

pure global buckling load ratio

firstly decreases then keeps almost constant

as the cut off region position

increases.

In the case of vibration, we define the

non-dimensional frequency

. A plate without cut off region, the

fundamental frequency for is found to be

2.42204. A square plate has arbitrary shape

of typical cut off region with the

coordinates

cut off

region position is occurred at = 0.3.

Results on the non-dimensional

fundamental frequency k versus are

shown in Figure 7. The result shows that

the decreases as the increases.

3.1.2 Clamped plate

For a clamped rectangular plate

without cut off region, the critical buckling

load is , this value should be the

maximum upper bound for all other

numerical results presented for a clamped

plate. A square plate has arbitrary shape of

typical cut off region with y-coordinates

y1=0.2b, y2=0.4b, y1=0.6b, y2=0.8b, (fixed

d1=0.3b and ζ1=η1=0.5b, ), cut off region

position is occurred at = 0.3. Results

on the critical load normalized with respect

to versus c1/b ( c1=0.1b, with

x1=0.4b, x2=0.5b, x3=0.6b; c1=0.2b with

x1=0.3b, x2=0.5b, x3=0.7b; c1=0.3b with

x1=0.2b, x2=0.5b, x3=0.8b; c1=0.4b with

x1=0.1b, x2=0.5b, x3=0.9b) are shown in

Figure 8. The result show that normalized

pure global buckling load ratio

decreases as the cut off region size c1/b

increases.

A square plate has arbitrary shape of

typical cut off region with the coordinates

x1(0.2b, 0.2b), x2(0.6b, 0.2b), x3(0.6b, 0.8b),

x4(0.4b, 0.6b), x5(0.2b, 0.4b) i.e. c1=0.2b,

d1=0.3b, ζ1=0.4b, η1=0.5b, Results on the

pure global buckling load normalized with

respect to versus are

shown in Figure 9. The result show that

normalized pure global buckling load ratio

firstly decreases and then keeps

almost constant as the cut off region

position increases.

A plate without cut off region, the

fundamental frequency for k is found to be

3.20406. A square plate having arbitrary

shape of typical cut off region with the

coordinates x1(0.2b, 0.2b), x2(0.6b, 0.2b),

x3(0.6b, 0.8b), x4(0.4b, 0.6b), x5(0.2b, 0.4b)

i.e. c1=0.2b, d1=0.3b, ζ1=0.4b, η1=0.5b, cut

off region position is occurred at =

0.3. Results on the non-dimensional

fundamental frequency k versus are

shown in Figure 10. The results show that

the k decreases as the increases.

3.2 Circular cut off region case

For a square plate has a circular cut

off region centered at and

cut off region radius r=0.3b that is shown

in Figure 11. We partition the circular cut

off region into rectangular shape of twelve

sub-regions and right triangular shape of

twelve sub-regions with coordinates X1(x1,

y1) to X12(x12, y12) corresponding to x- and

y-coordinates

where and n is integer, that is

shown in Figure 12. Some numerical

results are presented for four sides simply-

supported plate as well as clamped plate

under the global buckling and vibration.

3.2.1 Simply supported plate

Results on the pure global buckling

load normalized with respect to

versus were shown in Figure 13. The

result show that normalized pure global

buckling load ratio decreases as the

cut off region position increases.

Results on the non-dimensional

fundamental frequency k versus were

shown in Figure 14. The results show that

the decreases as the increases.

3.2.2 Clamped plate

Result on the pure global buckling

load normalized with respect to

versus was shown in Figure 15. The

result show that normalized pure global

buckling load ratio decreases as the

cut off region position increases.

Result on the non-dimensional fundamental

frequency k versus was shown in

Figure 16. The results show that the k

decreases as the increases.

3.3 Elliptical cut off region case

For a square plate has an elliptic cut

off region centered at and

cut off region length aspect ratio

as shown in Figure 17.

Similarly, we partition the elliptic cut off

region into rectangular shape of twelve

sub-regions and right triangular shape of

twelve sub-regions with coordinates

to corresponding to

x- and y- coordinates ,

where and n is

integer. Some numerical results are

presented for four sides simply supported

plate as well as clamped plate under the

pure global buckling and vibration.

3.3.1 Simply supported plate

Results on the pure global buckling

load normalized with respect to =3.0396

versus for for R=0.5~1.0 was shown

in Figure 18. The result show that

normalized pure global buckling load ratio

decreases as the cut off region

position increases. The non-

dimensional fundamental frequency k

versus was shown

in Figure 19. The result show that the k

decreases as the increases.

3.3.2 Clamped plate

Results on the pure global buckling

load normalized with respect to =6.9477

versus for R=0.5~1.0 was shown in

Figure 20. The result show that normalized

pure global buckling load ratio

decreases as the cut off region position

increases. The non-dimensional

fundamental frequency k versus for

for R=0.5~1.0 was shown in Figure 21. The

result show that the k decreases as the

increases.

4. ConclusionsThe pure global buckling and

vibration predictions due to reduced

flexural stiffness effect for four sides

simply supported as well as clamped

anisotropic laminates having arbitrary

shape of cut off region have been studied

by treating the cut off region with reduced

stiffness. The stiffness variation of the plate

is represented by Fourier series. The

numerical results are obtained by the Ritz

method of energy approach for plates

having typical arbitrary shape of cut off

region and utilizing the partition technique

under biaxial compression loads. We find

that the normalized pure global buckling

load ratio decreases as the cut off

region size increases. The non-dimensional

fundamental frequency k decreases as the

increases.

References1. Timoshenko, S. P. and Gere, J. M.,

"Theory of Elastic Stability," McGrow-

Hill Book Company, 1981

2. Wang, S. S., Zahlan N. M. and Suemasu

H., "Compressive Stability of

Delaminated Random Short-Fiber

Composites, Part I: Modeling and

Methods of Analysis," Journal of

Composite Materials, Vol.19, pp. 296-

316, 1985

3. Wang, S. S., Zahlan N. M. and Suemasu

H., "Compressive Stability of

Delaminated Random Short-Fiber

Composites, Part II: Experimental and

Analytical Results," Journal of

Composite Materials, Vol.19, pp. 317-

333, 1985

4. Simitses, G. J., Sallam, S. and Yin, W.

L., "Effect of Delamination of Axially

Loaded Homogeneous Laminated

Plates," AIAA Journal, Vol.23, No.9,

pp.1437-1444, 1985

5. Wang, J. T.-S., Lin, C. C. and Ong, C. L.,

"Analysis of Delaminated Composite

Structures II," Project Report No. NSC

83-0401-0-005-001, Dept. of Applied

Math., National Chung Hsing

University, Taiwan, 1994

6. Jane, K. C., and Hong, C. C., "Buckling

and Vibration of Rectangular Laminates

with Cut Off Regions," Mechanics

Research Communications, Vol. 27,

No. 1, pp. 101-108, 2000

7. Whitney, J. M., "Structural Analysis of

Laminated Anisotropic Plates,"

Technomic Publishing Company, Inc.,

1987

8. Szilard, R., "Theory and Analysis of

Plates Classical and Numerical

Methods," Prentice-Hall, Inc.,

Englewood Cliffs. New Jersey, 1974

Granger

(1)

(2)

(3)

(4) Granger 10

Granger

A Study on the Dynamic Relationships about

Government Bond Market in Taiwan, Japan,

United Kingdom & United States

Abstract

This article investigates the interrelationships of government bond market among

Taiwan, Japan, United Kingdom(UK), and the United States (US). Using cointegration,

vector error correction model (VECM), impulse response and variance decomposition

techniques to analyze this issue, we get the results as following (1)Except for Taiwan and

US, there are long cointegration trends for Japan and UK. (2) The VECM reveals that the

movement of current yield of Taiwan is independent from other countries. Both Taiwan and

US affect the movements of current yields of Japan and UK. Therefore, US have significant

effects on the other countries beside Taiwan. This result shows that United States still play an

important role in the whole world financial market. (3) As for the analysis of impulse

response, changing one standard deviation of each variable has the largest impulse at the first

period, and then gradually decreases. Accumulative effects of each market are positive.

Finally, analysis of variance decomposition shows that each variable has greatest

interpretative ability for itself. And the lead-lag relationship among four countries is that US

is higher than UK, Japan, and Taiwan. (4) The result in Granger causality tests is unapparent.

Only in 10% significant level, there are one-way causal relationship from Taiwan to Japan

and US. As well as there are also feedback relations among Japan, UK and US.

Key words: Cointegration, Vector Error Correction Model, Granger Causality Tests, Impulse

response, Variance Decomposition.

Ching-Jun Hsu : Associate Professor, Institute of Financial Management, Nan Hua UniversityYen-Hao Chen : Graduate Student, Institute of Financial Management, Nan Hua University

80

( 1991

Karfakis & Phipps 1999 ,

2000 )

Clare,Mara & Thomas(1995) Smith

(2002)

Smith (2002)

Phylaktis(1999)

LIBOR (

)

(TW)

(JP) (UK) (US)

Hendershott 1967

Lin and Swanson(1993) Engle

and Granger(1987)

Granger

domestic market

offshore market

1984 1989

(segmented Lin and

Leu(1994) Johansen

Granger

offshore market

perfectly linked

Lin and Swanson(1997)

simple autoregression models

Phylaktis 1999

Johansen Granger

multivariate Granger

causality tests

1970 1993

closely linked

(2000)

Johansen Granger

- 90

-

Clare,Mara & Thomas(1995)

Engle & Granger (1987)

1978-1990

Smith(2002)

Wilcoxon Rank Test

Johansen

Wilcoxon

Rank Test

3

(1995)

(comovement)

S & P 5 0 0

NIKKEI225

Johansen

Engle & Granger

Johansen

(1995)

Johansen

(Autoregressive Conditional

Heteroscedasticity ARCH)

(volatility)

ARCH

Granger &

Newbold(1974)

(Spurious Regression)

Granger

1.

ADF (Augmented Dickey-Fuller

1979) PP(Phillips-Perron 1988)

ADF PP

(H0)

ADF

Phillips-Perron

PP

PP

ADF

t p

A I C ( A k a i k e

Information Criterion)

p =0

PP

t

T

2. Engle & Granger(1987)

Engle & Granger

Johansen (1)

Engle & Granger

Xt

Yt I(1)

(OLS)

Xt Yt

β Xt Zt

Z t

ADF PP Zt

(4)

H0 =0

Zt Xt Yt

MacKinnon(1991) (2)

Johansen Xt N

1 I ( 1 )

(VAR)

ε t i i d

(Gaussian Pocess) 0

L (lag operator)

=1-L (5)

(6)

X t

(long-run

impact matrix)

(rank)

(full ranl)

I(1) r

Xt (5)

(VAR) r

N r

Johansen (1988)

(likelihood ratio

statistics) r

(trace test) H0

r H1 r+1

T

(maximum

eigenvalue test) H0 r

H1 r+1

H0

N r+1

3. Granger

Granger (1969)

X

Y X

Y

X Y

-

Granger

F (9) (10)

H0

H'0 Xt Yt

H'0

H0 Xt Yt

H0 H'0

Yt Xt

X t Y t

(feedback)

4. Engle & Granger

Granger Sims

Engle & Granger(1987)

bi ci d ECXt-1

t

Y Xt

Yt Yt Xt

VA R ( v e c t o r - a u t o r e g r e s s i o n

system)

5.

VAR

( )

VAR m

m2

(variance of forecast error)

innovation

innovation

(TW) (JP) (UK) (US)

TEJ

(I.F.S.)

1995 1 2003 2

98

1

1 1

1.6980

5% 6%

2

(C.V)

0.3246

ADF

PP

ADF PP

2

t

(white noise process)

AIC

2

t 1%

ADF PP

t

I(1)

Johansen(1990)

AIC

r = 0

r k

k Johansen

Trace test max test

Johansen & Juselius(1990)

( max test)

max test

3

Trace test

max test 1%

r = 0 r

1

max test

1 2

max test

Engle Granger(1990)

AIC(Akaike Information Criterion)

AIC

4

VECM AIC 2

2

6

(white noise)

Liung-Box

LM

Jarque-Bara

5

LB LM(1) LM(4) P 0.01

6

6

t

t -8.00985

1% -

0.93528

t 1.95988 5%

0.033833

Granger

Granger

Granger

F P-value

5% 1%

7

5%

10%

Granger

(VAR)

7.1

3 8

7.2

VEC

innovation

9 10 11

12

93.24%

89.14% 88.49%

52.95%

30.43% 11.74%

4.88%

0 %

2.85%

5.7% 13.65%

0% 2.78%

3.31% 8.44%

0% 3.2%

0.97% 8.5%

Granger

1 2

( )

Granger 5%

10%

( )

( )

( GARCH )

[1]

80

[2]

89

[3]

85

[4]

85

[5]

89 67 1-33

[6]Clare, A. D., Maras, M. and Thomas, S.

H., "The Integration and Efficiency of

International Bond Markets." Journal of

Business Finance & Accounting, 1995,

vol.22 (2), p313-322.

[7]Engle, R. E. and C. W. J. Granger,

"Cointegration and Error Correction:

Representation, Estimation, and

Testing." Econometrica, 1987, vol.55,

p251-276.

[8]Hendershott, P.H., "The Structure of

International Interest Rates: the U.S.

Treasury Bill Rate and the Eurodollar

Deposit Rate." Journal of Finance, 1967,

vol.22, p455-465.

[9]Johansen, S., "Estimation and

Hypothesis Testing of Cointegration

Vectors in Gaussian Vector

Autoregressive Models." Econometrica,

1991, vol.59, p1551-1580

[10]Johansen, S. and K. Juselius,

"Maximum Likelihood Estimation and

Inference on Cointegration with

Application to the Demand for Money."

Oxford Bulletin of Economics and

Statistic, 1990, vol.52, p169-209.

[11]Kenneth, L. Smith, "Government Bond

Market Seasonality Diversification, and

Cointegration: International Evidence."

The Journal of Financial Research,

2002, vol.25, p203-221.

[12]Karfakis, Costa and Anthony Phipps,

"Modeling the Australian Dollar-US

Dollar Exchange Rate Using

Cointegration Techniques." Review of

International Economics, 1999, vol.7

(2), p265-279.

[13]Lin, A., and P. E. Swanson, "Measuring

Global Money Market Interralationships:

An Investigation of Five Major World

Currencies." Journal of Banking and

Finance, 1993, vol.17, p609-628.

[14]Lin, A., and S. Leu, "Offshore Money

Markets Integration- Evidence of the

U.S. Dollar Yields in Taiwan Singapore,

and United Kingdom." Sun Yat-sen

Management Review, 1994, vol.2, p1-13.

[15]Lin, A., and P. E. Swanson, "The U.S.

Dollar in Global Money Markets: A

Multivariate Cointegration Analysis."

The Quarterly Review of Economics and

Finance, 1997, vol.37, p139-150.

[16]Phylaktis, K., "Capital Market

Integration in the Pacific Basin region:

An Impulse Response Analysis." Journal

of International Money and Finance,

1999, vol.18, p267-287.

(BJT)

R2

C

C

V(C)

BJT peak voltage detector

Abstract

This paper introduces a newly designed peak voltage detector, which consists of a

reference current generator, a differential amplifier with one-sided load transistor, a charge

transistor, a compensation current generator, a capacitor, a resistor, and an output stage.

Among them, the current in the reference current generator is copied in order to provide it to

the differential amplifier, the compensation current generator, and the output stage. The

differential amplifier serves as a comparator, and the charge transistor supplies the capacitor

with needed charge current. The compensation current generator is configured to provide

compensation current for compensating the voltage drop of the capacitor due to the base

currents in the bipolar transistors, and the output stage is configured to shift the voltage

signal on the capacitor to provide a precise peak voltage of input signal. The peak voltage

detector in this paper can accurately measure the peak voltage of input signal and it also

comes up with advantages like simple circuit design, minimal chip size, and good for use

with small devices. In addition, the inclusion of output stage can further prevent the held

peak voltage disruption from the accessing activities of the outer circuits. In the meantime,

the proposed peak voltage detector can further give a good elimination for the overshoot

voltage in the differential amplifier.

Key words: peak voltage detector, differential amplifier, overshoot voltage.

Ming-chuen Shiau: Associate Professor, Department of Electrical Engineering, HIT

(A/D converter)

(maximum likelihood decoding system)

[1]-[11]

[1] OP1

OP2 D1 D2

R1 R2 C

OrCAD PSpice

V(OUT) 0.01V

OP1

V(IN)

V(OUT) D1

C1

V(OUT) V(IN)

V(IN) V(OUT)

D2 D1

C

V(OUT) V(IN)

[12]-[13] [12]-[13]

[12]-[13]

OverShoot Voltage

Vos

(voltage transfer characteristic)

Vos

Vos

MOS

(BJT)

(BJT)

( R1

NPN MN1 )

(

NPN MN2 MN3 MN4

PNP MP1 )

PNP MP2

NPN MN5 MN6

PNP MP3 MP4

R2

NPN MN7 MN8

[12]-[13]

R2

SPICE

SPICE

V(IN) NPN

MN3 Vb(MN3) NPN

MN2 Ic MN2

NPN MN3 Ic

MN3

IR 1

PNP MP1 MP2

C

NPN MN3 Vb(MN3)

C V(C)

Ib(MN3) NPN MN3

R2

N P N M N 3

Vb(MN3) V(IN)

Vpeak

C

C V(C)

NPN MN3

Vb(MN3) Vpeak

OverShoot Voltage Vos

NPN MN2

NPN MN2

C

V(C)

Vos [14]

T 27 C

Vos 77.6mV

V IN

Vpeak NPN

MN2

C

V(C)

(6)

V(C) NPN

MN7 Vbe

V(OUT)

5 NPN

MN7 IR

NPN MN7 Vbe

IS

(saturation current) SPICE

[15]

(6) (7) (9)

V(OUT)

Vpeak

NPN MN2

(MN3) NPN MN3

(common -emitter

current gain) (12)

(11) (14)

(BJT)

Vpeak

Vpeak C

NPN MN3 Ib(MN3)

NPN MN7 Ib(MN7)

V(OUT)

V(OUT)

( PNP MP3 )

Ib(MN3) Ib(MN7)

NPN MN2

NPN

MN3 IR

NPN MN7

IR (

PNP MP3 )

IR NPN

MN3 (MN3)

IR NPN MN7

(MN7)

PNP MP4

IR NPN

MN5 (MN5)

PNP MP4 PNP MP3

NPN

2

NPN

Q2N2222 NPN PNP

Q2N3906 PNP

( PNP MP3 IS

) R1 R2

50K 650K

C 3nF

(BJT)

(1)

(2)

4 PNP

8 NPN 2 1

(3) V(OUT)

C

[1] Robert , F. C., and Frederick ,F. D.,Operational Amplifier & Linear

Integrated Circuits, Prentice-Hall,Englewood Cliffs, pp. 180-182, 1991.

[2] David,C.D.,"Tracking Peak Detect or,"U.S. pat. 5304939, Apr.1994.

[3] Ericson, M. N., and Simpson, M. L.,"ALow-power CMOS Peak Detect andHold Circuit for Nuclear," IEEE

Transactions on Nuclear Science,vol.42, pp.724-728 ,1995.

[4] Eiji ,S.,Kiyoshi, F., and Masafumi, K.,"Peak Detector," U.S. pat. 5546027,Aug., 1996.

[5] Ozguc ,I.H. , "Dual Stage Differ- entialAdaptive Peak Detector for DataCommunications Receivers," U.S.pat.5502746, Mar., 1996.

[6] Smith ,M.D., "Differential CrossCoupled Peak Detector," U.S. pat.5828240,Oct., 1998.

[7] Assadian, K., and Kosiec, J. H., "PeakDetector Circuit," U.S. pat.

5969545,Oct., 1999.[8] Lee ,J.C., and Brauns, G..T., "Offset-

compensated Peak Detector withOutput Buffering,"U.S. pat. 6051998,Apr., 2000.

[9] Wight ,M.S., Brazeau, S. H., and Grant,I. I., "Low Amplitude Peak Detector,"U.S. pat. 6064238, May, 2000.

[10]Chen, C.M., and Chen, P. F., "PeakDetector," U.S. pat. 6472861,Oct.,2002.

[11]

476418 2002[12]

5 1 7 1 6 12003

[13]

523592 2003[14]Laker, K. R., and Sansen,W. M. C.,

Design of Analog Integrated Circuits

and Systems, New York, McGraw-Hill,pp. 357-375,1994.

[15]Fjeldly, T. A., Ytterdal, T., and Shur,M.,Introduction to Device Modeling

and Circuit Simulation, Newyork,John Wiley & Sons, pp.166-185,1997.

1960 70 80 90

value chain

90

Research on Stone Industrial Development in

Taiwan

Abstract

In 1960s, the development of Taiwan stone industrial started in Hualien. It was one of

the essential processing industrials in the eastern of Taiwan. After rapid expansion from 70s

to 80s, stone industrial reached the peak of evolution in 90s. The processing facilities and

producing capacity of Taiwan, only comparatively lower than Italy, became the 2nd status in

the world. With the regression and fluctuation of Taiwan construction industrial, as a role of

material suppliers, the stone industrial encountered enormous market disadvantages.

This research presents essential issues of stone industrial, inclusive of industrial

progression, the analysis of success factors and difficulties, the operation strategies

suggested, and the following research on related topics, hopefully providing the constructive

suggestions for factory owners and the authority concerned.

Key words: stone industrial construction processing and manufacturing.

Lin Chen-Ju: Lecturer, Department of International Tourism Business, Taiwan Hospitality and TourismCollege

1960

70 80

90

STONE 1996 90

value chain position

90

90

Key Success Factors

SWOT

dimension stone

block

slab

( ) construction

materials

60

m o s a i c

( ) memorials

and landmarks

( ) furniture

and decoration

( ) souvenirs

1

ASTM C119-74

v a l u e

chain 1

( )

( )

( )

( )

( )

( )

1961

2

1

first mover advantage

followers

1967

1972

1976

1986

1987

1990 7

1992

1995

1996

( )

1990-1996

66%

14

10 10

( )

1980

( ) 85

( )

high dependence on oversea

suppliers

2

34

24

1 9

14

9

75 11

1980

93 7

1990

25 75

10-30

7-14

2-7

( )

( )

( )

300

Key Success Factors

1990

1930

80

75

Porter1980

T h r e e

Generic Strategies

stuck in the middle

2

Porter

SWOT

Porter1990

[U1] elements1 demand 2

logistics 3

strategies 4

strength weakness

opportunities threats

4

4

1

23

4

3

81

( )

5

5

( )

1.

87

technology transfer

2.

6

4

( )

1.

2.

3.

4. DIY

( ) DIY do it yourself

1. DIY

DIY

2.

( ) substitute goods

1.

2. Reverse

Engineering

3. Cost-Benefit

Analysis

4. Price

Assessment

5. Research of

Product Lines

6. Marketing

Channels

7.

( )

1.

2.

2 Porter1990

7

key success factors

( )

( )

( )

( )

( )

1.

2.

3.

...

4.

5.

.

1.

various issues.

2.

85

3.

85

4.

91

5.

85-90

6.

82 6

7. ,

8 5 1 2

.

1. Porter, M. E., Competitive Strategy:

Techniques for Analyzing Industries

and Competitors, New York: Free

Press, 1980.

2. Porter, M. E., Competitive

Advantage: Creating and Sustaining

Superior Performance, New York:

Free Press, 1985.

3. Stone World, various issues.

WTO

WTO

GATT WTO

A Study on the Effect of Entering WTO

on the Automobile Industry Policy in Taiwan

Abstract

The purpose of this study is to investigate the effect of entering WTO on the automobile

industry policy in Taiwan. First of all, the study analyzed the capital and subsidiary policy

on the automobile industry in Taiwan before entering WTO. After that, the study analyzed

the recent effect of WTO on the automobile industry. Finally, the study discussed the

advantages and disadvantages of WTO to Taiwan's automobile industry policy; moreover,

the recent situation was also explained in the study.

Key words: Policy, GATT, WTO, Tariff Quota.

Hua-Yin Liu: Postgraduate student, Graduate Institute of International Politics, National Chung HsingUniversity

WTO

GATT 1995 WTO GATT

GATT

WTO

WTO

WTO

WTO

2002 11

8 7 1

WTO

1953-1984

1985-1991

1991 2 1989

1 2002 12 23http://tw.news.yahoo.com/2002/12/23/finance/ctnews/3717907.html

2 2001http://www.motorsafety.com.tw/carsafety/%A5x%C6W%A8T%A8%AE%B5o%AEi%BE%FA%B5%7B.asp

80

1990

3

1983

303

1986

(Laser) 1500 1998

IMPREZA

2000

4

40

10

5 10

40

1953

GATT1994

1964

3 WTO61.222.52.195/net/Chin_zin/index.asp 2002 11 29

4 5 2 3 1996

http://www.moea.gov.tw/~ecobook/season/sa635.htm

6

6

2

4 1 2

10

6

WTO

GATT1994 16 1

3 3

GATT1994 3

1

1980

1985

WTO

6 2002 http://mail.nhu.edu.tw/~society/e-j/25/25-14.htm

1

WTO

1.

1996

1997

1997

1966

1977

1997

3000cc

1994

11

WTO 1994

2001

2.

GATT1994 3

1

3

GATT

3.

TRIM

TRIM

1991

4.

GATT1994

3 16

1 7

4

7 2001 10 12

WTO

2.

12

2

4

3%

3.

12%

WTO

9

1.

5

12%

1.

2.

GATT1994 11

3.GATT1994 20 2

4.

5.

WTO

6

7

2011

17.58

8 3 2001

1.

10,000

20

1990 1993

1

2 20

2.

7,500 4

9

WTO

140

9 WTO1999

http://www.moeaboft.gov.tw/global_org/wto/WTO-into/into6/into_001.htm

GATT1994 8

1

WTO

55

8

12

37

1.

2 0 0 0

25 2001

35

6 30

1

cc

2.

12 2

3%

3. 3 2

2000cc

2001cc 2

4.

12%

12%

2003 10 14

2004 1 1

2001 10 31

2011

2003

2003 1

WTO

WTO

2003 1

1 12 31 10

2002

2003

GATT1994 3

1

11

GATT1994

2001

11 21 7

15

10 WTO http://news.chinatimes.com/ 200210 23

11 2002 1http://www.cnfi.org.tw/cnfi/What's9101.htm

WTO

WTO

10

12

12

1999

13

NISSAN

14

9

12 http://ec.chinatimes.com.tw/scripts/chinatimes/iscstext.exe?DB=ChinaTimes&Function=ListDoc&From=61&Single=1 2002 9 6

13 2002 11 16

2001

1

1/3

2.

3 . 1

1 6 0 , 0 0 0

93 98

15,000

1

2 0 , 0 0 0

20,000

7,000 600

4.

7 815

150

1.

14 2001 12 3http://magazines.sina.com.tw/bnext/contents/20011105/20011105-006_5.html

15 WTO2001 12 5 http://www.epochtimes.com/b5/1/11/16/n149841.htm

16

1

29

2011 17.5

2000cc

25 2001cc

3601cc

1

3601cc

60 35

1 2007

35 30

3600cc

WTO 1

2002 8

1417

2002 1 9

35,723 100

1,025 2.9

2 0 , 0 8 9 5 7 . 0

534 1.5

W T O 1 3 , 6 2 5

38.6 18

2002

86.5

10

2003 1 9

16 WTO 1998 4 5http://www.chinabiz.org.tw/maz/InvCina/199804-050/199804-074-2.htm

17 2002 9 4http://ec.chinatimes.com.tw/scripts/chinatimes/iscstext.exe?DB=ChinaTimes&Function=ListDoc&From=80&Single=1

18

276,365

23,445 19

20

21

WTO

WTO

10

7

2002

86.5 22

19 http://www.ttvma.org.tw/chinesestatistic.htm

20 WTO2002 140-141

21 WTO 2003 1 2

22 2003 9 5 www.chb.com.tw/news/html/industry_report/09209/industrial_920905_3.html

1999

4,211

5.63%

42 3

WTO

WTO

WTO

24

25

2002

33.4

23 2001 4 24

http://www.chvv.ncue.edu.tw/vehicletec/vehicle%20technology%20educate/page001.htmwww.chvv.ncue.edu.tw/24 2002 9 625 ( ) http://www.cnfi.org.tw/

( 60

45

)

3

1 26

WTO

3

WTO

1989

WTO

WTO

40

26 2002 12 4

WTO

WTO

GATT

WTO

WTO

27

WTO

1.

2.

3. 4.

5.

5

4

27 2001 6

2002 3

2002 7

16

28

28 7http://www.moeaidb.gov.tw/idy/method/important/index.htm

92 94

2003 6 20

92 94 29

9

6

WTO

30 WTO

WTO

5

100% 31

NISSAN

29 92 94 2003 6 2030 -- 1998

http://216.239.33.100/search?q=cache:2itxogmRAwUC:www.chb.com.tw/ccbii/html/what_s_new_industrial_news_1202.htm+%E8%A3%95%E9%9A%86%E6%B1%BD%E8%BB%8A++%E8%A3%9C%E8%B2%BC&hl=zh-TW&ie=UTF-8

31 2002 9 27http://archive.udn.com/2002/9/27/NEWS/FINANCE/TRADE/1007650.shtml

WTO

WTO

WTO

1. 1999

1. 2002 WTO

1

135-162

2. 1998 WTO

50 221-256

3. 1999

WTO

http://www.moeaboft.gov.tw/global_or

g/wto/WTO-into/into6/into_001.htm

4. 2002

( )

http://mail.nhu.edu.tw/~society/e-j/25/25-

14.htm

5. 1996 9

2 3 http://www.moea.gov.tw/

~ecobook/season/sa635.htm

1 . 2002 9 4

http://ec.chinatimes.com.tw/scripts/chinat

imes/iscstext.exe?DB=ChinaTimes&Func

tion=ListDoc&From=80&Single=1

2. 2003 1 2 WTO

3. 2002 12 23

http://tw.news.yahoo.com/2002

/12/23/finance/ctnews/3717907.html

4. 2002 9 6

http://ec.chinatimes.com.tw/scripts

/chinatimes/iscstext.exe?DB=ChinaTimes

&Function=ListDoc&From=61&Single=

1

5. 2002 11 16

6. 2002 9 27

http://archive.udn.com/2002/9/27/

NEWS/FINANCE/TRADE/1007650.shtml

7. 2002 12 4

8. 2001 10 12

9. 2002 10 23

WTO

http://news.chinatimes.com/

1. 2001

2. 2003

92 94

3.

7http://www.moeaidb.gov.tw/idy/method/i

mportant/index.htm

1. 1998

http://216.239.33.100/search?q=cache:2i

txogmRAwUC:www.chb.com.tw/ccbii/htm

l/what_s_new_industrial_news_1202.htm

+%E8%A3%95%E9%9A%86%E6%B1

%BD%E8%BB%8A++%E8%A3%9C%E

8%B2%BC&hl=zh-TW&ie=UTF-8

2.

http://www.ttvma.org.tw/chinesestatistic.h

tm

3 .

2002 9 27 www.chb.com.

tw/news/html/industry_report/09/09109_i

mported_automobiles.html

4 .

2003 9 5 www.chb.com.tw/

news/html/industry_report/09/09209_imp

orted_automobiles.html

5. ( )

http://www.cnfi.org.tw/

6. 2001

http://www.motorsafety.com.tw/carsafety/

%A5x%C6W%A8T%A8%AE%B5o%AEi

%BE%FA%B5%7B.asp

7. 2002 1

19 2002

http://www.docamof.gov.tw/Frame-

7.htm

8. 2001 12 5

WTO

1

3

http://www.epochtimes.com/b5/1/11/16/n

149841.htm

9. 2002 WTO

61.222.52.195/net/Chin_zin/index.asp

10. 2001

http://www.chvv.ncue.edu.tw/

vehicletec/vehicle%20technology%20ed

ucate/page001.htmwww.chvv.ncue.edu.t

w/

11. 2002 1

http://www.cnfi.org.tw/cnfi/What's9101.

htm

12. 2001 12 3

http://magazines.sina.com.tw/bnext/cont

ents/20011105/20011105-006_5.html

13. 1998

WTO

http://www.chinabiz.org.tw/maz/InvCina

/199804-050/199804-074-2.htm

k-ε RNG

RNG k-ε

u(t)=0.6+0.6sin(50t) (Limit

Cycle)

Chaotic Character Phenomenon and

Numerical Analysis of Axisymmetric

Sudden Expansion Flow

Abstract

In this study, the standard k-ε turbulence model and the modified RNG- method-derived

k-ε turbulence model were employed in an axisymmetric sudden expansion flow with or

without swirling effect. A comparison of the velocity profiles was made with these

turbulence models. From the calculating results, the performance of RNG model is better and

in agreement with experiment than that of k-ε model. Furthermore, we innovatively utilized

periodical inlet conditions to simulate the sudden expansion flow. Surprisingly, striking

features of limit cycles of chaos were depicted in the phase diagrams. Thus, we could have

an idea that the intermittent sudden expansion flow is both turbulent and chaotic character for

the fluid/system dynamics. In brief, the RNG turbulent model could characterize the flow-

field energy cascade and dissipation processes more accurately and efficiently.

Key words: Turbulence Model, Sudden Expansion, Limit Cycle, Chaotic.

Jiunn-Shean Chiang Instructor, Department of Mechanical Engineering, HIT.Yen-Hung Liu Graduate student, Department of Mechanical Engineering, NCHU.

( )

(Recalculation)

(Swirling Type) (Side Inlet)

(Flame Holder)

B a c k

Roschke

H. J. Sheen [1]

(

) P. A. Dellenback[2]

Baoyu[3] CFX K-

Epslion

(Chaos)

Chang[4] C

[5]

(LES)

Koronaki[6]

(wall function)

RNG

Charles[7]

k-ε

F. D. Stull[8] k-ε

k-ε

Rodi[9] (Eddy)

k-ε

P. Bradshaw[10]

Yakhot & Orszag [9][11]

RNG model k-ε

model

RNG model

1

Dellenback[2]

(1)

(Newtown Flow)

(2) (Ful ly

Develop)

(3) (Adiabatic)

(Impermeable)

3.1

X

Y

S

Launder Spalding[12]

k-ε

Yakhot Orszag[9][11]

(Mean Strain)

k-ε

[13] SRNG

(3.7)

3.2

(A) (Inlet Condition)

1.0

k ε

POLIS[13]

T I

5%

Dellenback[2]

Chang[4]

(3.8)

50 k

POLIS[13]

5%

(B)

(Neumann boundary condition)

Dellenback[2] Chang[4]

(C)

(No-Slip

Condition)

(D)

( A x i s -

symmetric Condition)

(Finite Volume Method)

(Staggered Grid System)

[14]

(Hybrid Scheme)

CFL

SIMPLEST

(Under-Relaxation)

10-4

5.1

2 4 (KE-

Uniform, RNG-Uniform)

(KE-Exp, RNG-Exp)

2 4

(Exp.Ref[2])

5.2

2

2

RNG k-ε

x/D=2.0 RNG

RNG

RNG k-ε

5 6

RNG k-ε

RNG Xr/h=9.1 k-ε Xr/h=8.2

Xr/h=9.3 RNG

Koronaki[6]

Yakhot[9]

RNG k-ε

x/D=2.0

RNG

(

) (

)

3 4 x/D=2.0

RNG

7

k -ε KE

RNG

(Dissipation)

k-ε RNG

RNG

(e.g. Komologov

Energy Spectrum) k-ε

RNG

k-ε

5 RNG C

5.3

5.3.1.

Baoyu[3] 8

k-ε RNG

(1)

( F r e e

Vortex)

(2)

9

RNG

5.3.2.

RNG model k-ε model

C RNG model

RNG

Chang[4]

9

(e. g. )

LES(Large Eddy

Simulation) DNS(direct Numerical

Simulation)

C

LES DNS

5.4

RNG

C F L

Courant number 0.1

10

2.0

11

T = 0.04 (f=

25Hz)

(resonance)

(lock-over) (lock-

in)

12 15

(Chaotic)

( S i m p l e

Attractor)

(Stranger Attractor)

-

(

)

k-ε RNG

(1)

(2)RNG model k-ε model

C RNG model

(3)

(4)

1.Sheen H. J., W. J. Chen, T. L. Huang,

"Correlation of Swirl Number for a

Radial-Type Swirl Generator",

Experiment Thermal and Fluid Science,

Vol. 12, pp. 444-451, 1969.

2.Dellenback P. A., D. E. Metzger, G. P.

Neitzel, "Measurements in Turbulent

Swirling Flow Through an Abrupt

Axisymmetric Expansion", AIAA

Journal, Vol. 26, No. 6, pp. 669-681,

1988.

3.Baoyu G., A. G Tim. Langrish, F. F

David., "Simulation of Turbulent Swirl

Flow in an Axisymmetric Sudden

Expan-

sion", AIAA Journal, Vol. 26, pp. 1-15,

2002.

4.Chang K. C., C. S. Chen, "Development

of A Hybrid Turbulence Model for

Swirling Recirculating Flows Under

Moderate to Strong Swirl Intensities"

Int. Journal for Numerical Methods in

Fluids, Vol. 16, pp. 421-443, 1993.

5.Jorg Schluter, "Influence of

Axisymmetric Assumptions on Large

Eddy Simulations of a Confined Jet and

a Swirl Flow", submitted for publication

to International Journal of

Computational Fluid Dynamics, 2001.

6.Koronaki E. D., H. H. Liakos, M. A.

Founti, N. C. Markatos, "Numerical

Study of Turbulent Diesel Flow in a

Pipe with Sudden Expansion", Applied

Mathematical Modelling, Vol. 25, pp.

319-333, 2001.

7.Pierce C. D., P. Moin, "Method for

Generating Equilibrium Swirling Inflow

Conditions", AIAA Journal, Vol. 36,

No. 7, pp. 1325-1327, 1998.

8.Stull F. D. and R.R. Craig, "Investigation

of Dump Combustors with Flame

Holders", AIAA-75-165, 1975.

9.Yakhot V., S. A. Orszag, "Development

of Turbulence Models for Shear Flow

by a Double Expansion Technique",

Phys. Fluids A, Vol. 4, pp. 1510-1520,

1992.

10.Bradshaw P., "The Analogy Between

Streamling Curvature and Buoyance in

Turbulent Shear Flow", J. Fluid Mech.,

Vol. 36, pp. 171-191, 1969.

11.Yakhot V., S. A. Orszag,

"Renormalization Group Analysis of

Turbulence. I. Basic Theory", Journal of

Scientific Computing, Vol. 1, No. 1, pp.

3-51, 1986.

12.Launder B. E., D. B. Spalding, "The

Numerical Computation of Turbulent

Flow", Computer Methods in Applied

Mechanics and Engineering, Vol. 3, pp.

269-289, 1974.

13.Concentration, Heat and Momentum

Limited (CHAM), London, POLIS:

PHOENICS On-Line Information

System, http://www.cham.co.uk, 2003.

14.Patankar S. V., "Numerical Heat

Transfer and Fluid Flow", Hemisphere

Publishing Co., New York, 1980.

" GPS/INS "

GPS INS

GPS/INS

(2nd order) (Guassian Noise)

(

) (Runge-Kutta method or Hermite interpolation)

GPS/INS

GPS/INS

/

/

Non-linear GPS/INS navigation for

autonomous mobile vehicles

Abstract

Nowadays, there has been a major upsurge of interest in the integrated GPS/INS usage

as a cost-effective way of providing the accurate and the reliable navigation for military and

civil applications. Therefore, in this research, we propose a concisely and well-designed non-

linear architecture for the autonomous mobile vehicles. This system evaluates performance

with a decentralized architecture for the fusion of information from different asynchronous

sources.

The GPS/INS filter mechanism, in this research, was developed with an error model

including linear and non-linear components. The latter consists of a quadratic function of

states and was approximated by a Gaussian-noise term thereby allowing Kalman filter being

used under the varying measurement sets. Moreover an integrator dealing with navigation

system has to provide continuous and dense outputs at the equidistant-time steps and also

supplemented by a Runge-Kutta method or Hermite interpolation scheme.

Conclusively, the corresponded network simulations and experimental results will be

presented to demonstrate this novel non-linear navigation architecture of integrated GPS/INS

in usage of the autonomous mobile and/or remote pilot vehicles.

Key words: Non-linear GPS/INS navigation for autonomous mobile vehicles.

Chih-Yeh King Assistant Professor, Department of Electrical Engineering, HIT

(GPS/INS)

(

) [1~3] US-

GPS [4,5] GLONASS

[6,7]

GPS/INS

GPS

INS

INS/GPS Kalman

Filter

GPS INS

(Pseudo-

range, Pseudo-range-rate) GPS/INS

GPS IMU

[8,9]

GPS

GPS INS

GPS/INS

GPS/INS

[10~15] Strapdown

(

) GPS

1 Hz.

(Kalman Filter) 5 Hz

GPS

GPS

(Second-

order terms)

Quadratic Filter

2.1 GPS/INS

(Body-axes)

x- (N) y-

(E) z- (D)

(State Model)

B o d y - t o - L o c a l

2.2

(a)

(b)

(c)

(Color-noise) Quadratic

Filter

(Color-noise) Quadratic

Filter

(Noise Correlated)

(e)Integrated nonlinear GPS/INS error

model

(d) Vehicle's rotation Equation

Predicator's criterira:

[16~18] [H]

GPS (Pseudo-range &

Pseudo-range rate)

GPS (dB, dF)

GSM/GPRS

GPS

P C (

)

( (a), (b))

[P(0|0)], [Q], [R]

( (c))

( (d))

(GPS/INS)

(Hermite Interpolation)

(On-board)

Kalman

([P(0|0)], [Q], [R]

)

1. [P(0|0)], [Q], [R]

2. [H]

3. GPS

LabVIEW MATLAB

GPS/INS

1. LabVIEW

( R S - 2 3 2 ,

PIMCIA, USB) GPS

NIMA-0183

(Pseudo-range &

Pseudo-range rate)

(WGS84/TWD97)

2. GPS/INS

MATLAB

(State Predicator)

0.2 sec 1

s e c (

Accelerometer, Gyroscope)

GPS/INS

3. GPS/INS

(Second-order forms)

Quadratic filter

Color noise

4. (

[1] E. M. Nebot and H. Durrant-Whyte, "A

high integrity navigation architecture for

an outdoor autonomous vehicles",

Robotics Autonomous Systems Vol. 26,

pp. 81-97, 1999.

[2] J. Hunter, W. Kosmalski, and P. Truong,

"Vehicle Navigation Using Differential

GPS", International Conference of IEEE

on Position Location and Navigation

System, pp. 392-398, 1990.

[3] R. C. Hart, C. J. Gramling, J. K.

Deutschmann, A. C. Long, D. H. Oza,

W. L. Steger, "Autonomous navigation

initiatives at NASA GSFC flight

dynamics division", Proc. IAS

International Astro-dynamics

Symposium, 96-C-23, pp. 125-130,

1996.

[4] Eckley, G.P., "Navstar Global

Positioning System Coordinate System

Definitions Representations and

Analysis", The Magnavox Company

Advanced Products Division Report

MRL R-5052, pp. 4-22, March 31, 1975.

[5] Martin, E.H., "Navstar Global

Positioning System Navigation

Mechanization Analysis", The

) G P S

(Pseudo-range & Pseudo-range rate)

Hermite Interpolation

Curve

Fitt ing

[19~24] (DSPs

TMS320C54x) /

GSM/GPRS

GPS/INS

Magnavox Company Advanced Products

Division Report MRL R-5070, pp. 17-

51, May 12, 1975.

[6] J. K. Ray, O. S. Salychev, M. F.

Cannon, "The modified wave estimator

as an alternative to a Kalman filter for

real-time GPS/GLONASS INS

integration", Journal of Geodesy, Vol.

73: pp. 568-576, 1999.

[7] R. E. Phelts, D. Akos, and P. Enge,

"SQM validation report for GNSSP Wg-

B Meeting", Dept. of Aero/Astron.

Engineering, Standford University, May

2000.

[8] M. A. Lewis, A. H. Fagg, G. A. Bekey,

"The USC autonomous flying vehicle:

An experience on real-time behavior

control", IEEE International Conference

on Robotics and Automation, pp. 422-

429, May 1993.

[9] T. S. Stombaugh and S. A. Shearer,

"The DGPS-based automatic guidance of

agriculture vehicles", Proc. European

Conference on Precision Agriculture, pp.

121-126, 2001.

[10] Wong, R.V.C., "Development of a

RLG Strapdown Inertial Survey System"

Department of Surveying Engineering

Calgary, pp. 31-68, Alberta, 1988.

[11] Piscane, V.L., Moore, R.C.,

"Fundamentals of Space Systems",

Oxford University Press, pp.245-304,

New York, NY, 1994.

[12] M. Bozorg, E. Nebot, and H. Durrant-

Whyte, "A decentralized navigation

architecture", Proc. IEEE-ICRA,

Belgium, pp. 3413-3418, 1998.

[13] F. A. Faruqi, "Non-linear mathematical

model for integrated globe position and

inertial navigation systems", Applied

Mathematics and Computation, Vol. 115,

pp. 191-212, (2000).

[14] H. D. Stevens, E. S. Miles, S. M.

Rock, R. H. Cannon, "Object-based task-

level control: A hierarchical control

architecture for remote operation of

space robots", Proc. AIAA/NASA

Conference on Intelligent Robotics in

Field, Factory, Service and Space, pp.

264-273, Houston, 1994.

[15] P. W. McBurney, "A robust approach

to reliable real-time Kalmam filtering",

Proc. Of IEEE Position, Location, and

Navigation Symposium, 549-556, March

1990.

[16] Handley, S., Langley, P. and Rauscher,

F., "Learning to predict the duration of

an automobile trip", Proc. AAAI, 4th

Annual Conference on Knowledge

Discovery and Data Mining, pp. 219-

223, New York, 1998.

[17] S. P. Karatsinides, "Enhancing filter

robustness in cascaded GPS-INS

integration", IEEE Trans. on Aerospace

& Electronic Systems, 1001-1008, Oct.

1994.

[18] C. C. Arcasoy and B. Koc, "Analytical

solution for continuous-time Kalman

tracking filters with colored

measurement noise in frequency

domain", IEEE Trans. on Aerospace &

Electronic Systems, 1059-1063, Oct.

1994.

[19] C.Y. King, "Virtual Instrumentation-

Based System in a Real-Time Telemetry

of GPS/GIS", Proceedings on IEEE

RAST2003, Ref: S6BV2, Nov. 20~22,

2003. (Accepted)

[20] "

" 2003

A2 47 ~ 52

Oct. 17, 2003.

[21] C.Y. King, "Study of a Concise

Programming for GPS Positioning and

Navigation", Proceedings on CAC 2003,

C10-4, pp. 548-553, March 13~14, 2003.

[22] C.Y. King, "Virtual Instrumentation-

Based System in Real-time Positioning

of GPS"

Paper ID: F-056, Taiwan, Nov.

14~16, 2002.

[23] C.Y. King, "Study on a real-time

LabVIEW prototype of the GPS

system", AMTE 2002 IEEE/ASME

International Conference, A103, Chia-yi,

Taiwan, Aug. 11~14, 2002.

[24] "

"

105 ~ 114

Aug. 09,

2002.