國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
143
CHAPTER 5. SUMMARY AND FUTURE STUDY
From the recent literatures review, we found that a producer takes the action of perfect repair for nonconforming items for instrument with measurement error, under the producer may or may not provide a refund to middleman and the middleman may provide total or partial refund to customer then the optimal specification limits, buying price and selling price are determined to maximizing a profit model of the middleman, and considering both the order quantity and specification limits in the profit model of middleman have not been discussed.
In this study, we first determined the optimal specification limits to maximize profit model for producer taking the action of perfect repair or action of sell low price for observed nonconforming item under instrument with measurement error. Secondly, we considered the producer constructed
economic 2
P YP
Y lnS
EWMA − EWMA
chart to control process quality with maximal profit per unit time but did not take complete inspection. From the results of comparing the profits of complete inspection and control process quality in a fixed observed time, we found that under the combinations of small R, high Cpr, high Csc, and high PPM, the profit of conducting process control is higher than conducting a complete inspection. Hence, process control should be favorable by producer; otherwise the producer should take complete inspection.The third, under the first situation we considered that a middleman bought products from the producer and sold products to a customer. Assume the specification limits of the producer and the middleman are the same, the middleman’s instrument may or may not exist measurement error, for returning nonconforming products the producer may or may not provide a refund to middleman and the middleman may provide total or partial refund to customer, and the buying price and selling price of products are decided by middleman. By maximizing a derived expected profit per item of the middleman, the optimal specification limits, buying price and selling price are determined. From the results of numerical analyses, we found that the larger specification limits and selling price the larger profit, but the lower buying price the larger profit; the expect profits per item are almost same for middleman provided total or partial refund to customer and the producer may or may not provide a refund to middleman; to have larger profit, middleman should buy products from producer took the
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
144
action of perfect repair when middleman’s instrument existed measurement error or middleman’s instrument existed no measurement error but producer’s instrument existed small measurement error.
The fourth, under the middleman provides a partial refund for customer and get the refund from producer we investigated whether the order quantity may affect the optimal specification limits, since the optimal specification limits are all the same ( or equal to the upper bound of specification
constraint) in the third section. However, from the numerical analyses we found that the order quantity do not affect the optimal specification limits either.
For the future study, we may extend the provided model to consider the producer and middleman with different tolerances, for a complete inspection plan may let the ratio of the expected in-control time to m unit time be a decision variable in the profit model, adaptive economic process control scheme to improve out-of-control detection speed, and also consider the widely popular consignment policy to determine the ordered quantity for middleman.
‧
[1] Bennett, G. K., Case, K. E., and Schmidt, J. W. (1974). The economic effects of inspection errors on attribute sampling plan. Naval Research Logistics Quarterly, 21, 431-464.
[2] Biegel, J. (1974). "Inspection errors and sampling plans". AIIE Transactions, 6(431-464).
[3] Bisgaard, S., Hunter, W. G., and Pallesen, L. (1984). "Economic selection of quality of manufactured product". Technometrics, 26, 9-18.
[4] Case, K. E. (1980). "The P control chart under inspection error". Journal of Quality Technology,
12, 1-9.
[5] Chen, C. H. (2008). "Joint determination of optimum process mean and economic specification limits for rectifying inspection plan with inspection error". Journal of the Chinese Institute of
Industrial Engineers, 25(5), 389-398
[6] Chen., and Chung, K. J. (1994). "Inspection error effects on economic selection of target value for a production process". European Journal of Operational Research, 79, 311-324.
[7] Chen., and Chung, K. J. (1996). "Selection of the optimal precision level and target value for a production process: the lower-specification-limit case". IIE Transactions, 28, 979-985.
[8] Crowder, S. V., and Hamilton, M. D. (1992). "An EWMA for monitoring a process standard devation". Journal of Quality Technology, 24, 12-21.
[9] Dooris, A. L. (1977). "The effect of inspection error on the operation of a c-chart". AIIE
Transactions, 9, 312-315.
[10] Duffuaa, S., and Siddiqui, A. W. (2003). "Process targeting with multi-class screening and measurement error". International Journal of Production Research, 41, 1373-1391.
[11] Duncan, A. (1956). "The economic design of X-bar chart used to maintain current control of aprocess". Journal of The American Statistical Association, 51, 228-242.
[12] Fathi, Y. (1990). "Producer-consumer tolerance". Journal of Quality Technology, 22, 138-145.
[13] Feng, and Kapur. (2006). Economic design of specifications for 100% inspection with imperfect measurement systems. Quality Technology & Quantitative Management, 3(2), 127-144.
[14] Ferrell, W. G., and Chhoker, A. (2002). "Design of economically optimal acceptance sampling plans with inspection error". Computers and Operations Research, 29, 1283-1300.
[15] Gan, F. F. (1995). "Joint monitoring of process mean and variance using exponentially weighted moving average control charts". Technometrics, 37, 446-453.
[16] Golhar, D. Y. (1987). "Determination of the best mean contents for a canning problem". Journal
of Quality Technology, 19, 82-84.
[17] Hunter, W. G., and Kartha, C. P. (1977). "Determining the most profitable target value for a production process". Journal of Quality Technology, 9, 176-181.
[18] Kapur, K. C. (1987). "Quality loss function and inspection". Quality Control, Innovations in
Quality: Concepts and Applications.
[19] Kapur, K. C., and Wang, D. J. (1987). "Economic design of specifications based on Taguchi’s concept of quality loss function". Proceeding of American Society of Mechanical Engineers.
[20] Lee, M. K., and Kim, G. S. (1994). "Determination of the optimal target values for a filling
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
146
process when inspection is based on a correlation variable". International Journal of Production
Economics, 37, 205-213.
[21] Lorenzen, T. J., and Vance, L. C. (1986). The economic design of control charts: a unified approach. Technometrics, 28(1), 3-10.
[22] Lucas, J. M., and Saccucci, M. S. (1990). "Average run lengths for exponentially weighted
moving averages schemes using the Markov Chain approach". Journal of Quality Technology, 22, 154-162.
[23] Maghsoodloo, S., and Li, M. (2000). "Optimal asymmetric tolerance design". IIE Transactions,
32, 1127-1137.
[24] Mei, W. H., Case, K. E., and Schmidt, J. W. (1975). "Bias and impercision in variable acceptance sampling: effects and compensation". International Journal of Production Research, 13, 327-340.
[25] Menzefricke, E. (1984). "The effects of variability in inspection error". Journal of Quality
Technology, 16, 131-135.
[26] Montgomery, D. C. (1985). Economic Design of X-bar Control Charts for Two Manufacturing Process Models. Naval Research Logistics Quarterly, 32, 631-646.
[27] Sheng (2012). "最小成本下, 規格及 X-bar S 管制圖之設計". Unpublished master thesis, Department of Statistics, NCCU.
[28] Stemann, D., and Weihs, C. (2001). "The EWMA-X-S control chart and its performance in the case of precise and imprecise data". Statistical Papers, 42, 207-223.
[29] Tang, K. (1987). "The effects of inspection error on a complete inspection plan". IIE Transactions,
19, 421-428.
[30] Tang, K. (1988). "Economic design of product specification for a complete inspection plan".
International Journal of Production Research, 26(203-217).
[31] Tang, K., and Lo, J. (1993). "Determination of the process mean when inspection is based on a correlated variable". IIE Transactions, 25, 66-72.
[32] Tang, K., and Schneider, H. (1990). "Cost effectiveness of using a correlated variable in a complete inspection plan when inspection error is present". Naval Research Logistics Quarterly,
37, 893-904.
[33] Yang, S.F. (2013). "Using a new VSI EWMA average loss control chart to monitor changes in the difference between the process mean and target and/or the process variability". Journal of
Mathematical Modelling. online publish.
‧
‧
Appendix B. Calculate the Average Run Length
We calculate the ARL’s forEWM AY and 2lnSYP
EWMA
individual by using the Markov chain approach (describe following), and byYP lnS2 ,we can get the ARL value for combing the
YP
EWMA
and 2lnSYP
EWMA
scheme The following will describe the ARL calculation procedure for theYP
For in-controlEWMAY control charts is
YP
UCL
) and lower control limits (LCL
EWMAYP) into g=2m+1 , the number of states , sub-intervals of width 2δ, whereYP YP
EWMA EWMA
UCL LCL
δ 2 g
= −
‧
interval. (Like the following figure)Step3 The statistic
( )
Step4 The transition probability matrix for the transient state ofYP[ ( )]
Expressed by Normal CDF For in-control
, then the transition probability matrix of
Y
which calculated by in-control normal distribution is‧
, then the transition probability
matrix ofZYwhich calculated by out-of-control normal distribution is
[ ( )] Step5 Assume that process begins from state 0 (zero state) ; thus
let
{ {
m terms m terms
( 0,...,0,1, 0,...,0)
zs
=
p
,wherepzsrepresent the zero state probability vector.Step6
Where I is the g×g dimension identity matrix , and 1 is the g×1 dimension vector with all components
EWMA
control charts is[ ( )] [ ( )]
‧
UCLEWMA ) and lower control limits
( lnS2
YPI
LCLEWMA ) into g=2m+1 , the number of states , sub-intervals of width 2δ, where lnSY2PI lnSY2PI
Step4 The transition probability matrix for the transient state of 2
YP
σ
is the standard deviation of 2YP
ln S
, andΦ(
⋅)
is the cdf of standard normal distribution.Expressed by Normal CDF
For in-control
( [ ( )] [ ( )])
ln S
which calculated by in-control normal distribution is[ ( )]
2
YPI t 1,t
I,lnS
= p
−jk j,k = − m,..., 1,0,1,...,m −
R
‧
lnS ~ N E ln S ,Var ln S
, then the transition probability matrix of 2YO
S
ln
which calculated by out-of-control normal distribution is[ ( )]
Step5 Assume that process begins from state 0 (zero state) ; thuslet
{ {
m terms m terms
( 0,...,0 ,1, 0,...,0 )
zs
=
p
,wherepzsrepresent the zero state probability vector.Step6
Where I is the g×g dimension identity matrix , and 1 is the g×1 dimension vector with all components.