Input control in a batch production system with lead times, due dates and random yields
|
|
- Victoria Lynch
- 5 years ago
- Views:
Transcription
1 European Journal of Operational Research 126 (2) 371±385 Theory and Methodology Input control in a batch production system with lead times, due dates and random yields Yunzeng Wang a, *, Yigal Gerchak b,1 a Department of Operations Research and Operations Management, Weatherhead School of Management, Case Western Reserve University, 19 Euclid Avenue, Cleveland, OH , USA b Department of Management Sciences, University of Waterloo, Waterloo, Ont., Canada N2L 3G1 Received 1 September 1998; accepted 1 April 1999 Abstract The presence of random yields in many manufacturing processes can considerably complicate production planning and control. We investigate a batch production system with due dates, where the yield of each batch is random and the production lead time is longer than the time interval between starting consecutive batches. To satisfy an order with a given due date, several input batches could be initiated. But the realized yields of batches still in process are unknown when the next batch size needs to be determined. We formulate this general problem as a dynamic program with the objective of minimizing the total expected discounted costs. For a simple version of the model with linear cost parameters and one work-in-process batch, we show that the structure of the optimal input control policy is of a single critical level type for the work-in-process batch size. We prove that, for given outstanding demand, this critical level becomes larger as the number of input opportunities becomes smaller. Furthermore, if the discount factor equals to 1, this critical level is shown to be strictly greater than the outstanding demand. Production often starts earlier than necessary in order to utilize yield realizations of initial batches for judicious choice of later batches, and to achieve diversi cation over time. Ó 2 Elsevier Science B.V. All rights reserved. Keywords: Lot sizing; Random yield; Production control; Due dates 1. Introduction Consider an imperfect production system, which processes batches sequentially. At each review epoch (beginning of a review period) a batch of items is input into the system for processing. The processing time, * Corresponding author. Tel.: ; fax: addresses: yxw36@po.cwru.edu (Y. Wang), ygerchak@engmail.uwaterloo.ca (Y. Gerchak). 1 This research was supported by the Natural Science and Engineering Research Council of Canada. Helpful comments by referees are acknowledged //$ - see front matter Ó 2 Elsevier Science B.V. All rights reserved. PII: S ( 9 9 ) 295-7
2 372 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371±385 or production lead time, for a batch is longer than the review period. Thus, at any time, several work-inprocess batches at di erent processing stages will be present in the system. Production processes of this type are quite common. They can be generically viewed as a ``conveyor'', which carries the batches through the system, as shown in Fig. 1(a), where I k is the size of batch k. Quality inspection takes place at the end of the line, and the net yield of good units is random. Now suppose that an order is received, and it is to be satis ed within a due date. Then, the number of input batches designated to satisfy the order is limited, due to production lead times. The problem is to choose the input quantities for each of these consecutive input batches in order to balance the potential costs of over production and shortage penalty at the due date. What really makes this problem challenging, both theoretically and practically, is the fact that the yields of batches that are in process are not known when the size of the next input batch needs to be selected. Fig. 1(b) provides an example to illustrate the temporal dynamics of the system, where the production of a batch takes two review periods. Demand due date is four review periods away. As a consequence, only three batches can be input to meet the due date. The batches are labeled in the reverse order of their input sequence. At the point when the size of the last batch I 1 is to be determined, the rst two input batches I 3 and I 2 are still in process and their yields (number of good items), namely, Y I3 and Y I2, respectively, are not yet known. Our model shares some common features with several branches of the literature on lot sizing with random yields (Yano and Lee, 1995). If specialized to the case of no work-in-process (i.e., up-to-date information about the realized yields of all previous batches), our physical scenario becomes similar to that of multiple-lot-sizing (``reject-allowance'') production-to-order (e.g., Beja, 1977; Sepehri et al., 1986; Grosfeld- Nir and Gerchak, 1996 and references therein). Within that literature, Klein (1966) and Sepehri et al. (1986) address the case of limited number of production runs, while White (1965), Grosfeld-Nir and Gerchak (199) and Wein (1992) consider stochastically proportional yields. But, the economic trade-o in this branch of literature is between setup and variable production costs, while our focus is on trading-o production and overage costs against penalties for not completing the order by its due date. The several ``stages'' of work-in-process in our setting are reminiscent of multi-stage production models with random yields (Lee and Yano, 1988; Barad and Braha, 1996; Grosfeld-Nir and Ronen, 1993; Grosfeld-Nir and Gerchak, 1998). Yet, in these models batches are inspected after every processing stage, and thus the whole decision structure, as well as the economic trade-o s, are quite di erent. Although ours is a single-product setting, the fact that lot sizing decisions have to be made before the results of previous (or simultaneous) decisions are known is somewhat reminiscent of lot sizing problems in Fig. 1. The batch production system.
3 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371± assembly systems with random yields (Singh et al., 199; Gerchak et al., 1994; Gurnani et al., 1996) as well as control problems with random lead times (Chu et al., 1993; Proth et al., 1997). Due-date-based models in production management typically ask when to start some activities, and in what order, so as to minimize tardiness and earliness penalties. Our random yield (rather than random processing time) setting raises lot sizing issues, rarely discussed in conjunction with due dates. Our model minimizes the total expected discounted costs of production, overage and penalties for the portion of an order not satis ed by the due date. No explicit earliness penalties for units completed before the due date are included. We rst formulate a very general model with an arbitrary number of work-inprocess batches, general cost functions and capacity constraints. We then present a detailed analysis for a simpli ed version of the general model. In this simpli ed model, we assume that all the cost functions are linear and that there is only one work-in-process batch at any decision point. We show that the optimal policy is of a single critical level type. That is, a new input batch is started if and only if the size of the workin-process batch is less than a critical level. We prove that, for given outstanding demand, this critical level becomes larger as the number of input opportunities becomes smaller. Furthermore, in the absence of discounting, this critical level for each input batch is shown to be strictly greater than the outstanding demand. The reasons for starting production earlier than needed to meet the due date is that information about realized yields of early batches helps in judicious choice of latter batch sizes, and the multiple trials provide yield diversi cation over time, similar to what has been observed by Henig and Gerchak (199) in a periodic-review setting with random yields. 2. Problem de nition and formulation The production system under consideration can be described as follows. At every review epoch one ``batch'' of items is input into the system for processing. The processing/production lead time of each batch takes a xed time equal to K review periods, independent of the batch size. As a consequence, there are always up-to K batches in process just before any additional batch is fed into the system at the beginning of a review period. Note that those batches are at di erent processing stages. Inspection, through which defects (bad items) will be picked out, takes place at the last ``processing'' stage; the yield (# of good items) from any batch is random. Speci cally, we model the random yield associated with a batch of size x by the stochastically proportional (multiplicative) yield model: Y x ˆ x U; where, U is the random fraction of good items, which is assumed to be independent of x. The random multiplier U has a density g(á) on [, 1], cumulative distribution G(á) and a mean l; and Y x the resulting yield (good items). We assume that the multipliers (U's) corresponding to di erent batches are i.i.d. This common-cause yield model has been used extensively by various researchers (e.g., Lee and Yano, 1988; Henig and Gerchak, 199; Wang and Gerchak, 1996; see also Yano and Lee (1995) and references therein). Suppose we are working towards lling a single order with a due date of H H > K periods away. Since the last batch should be fed at least K periods (the production lead time) before this due date, the maximal number of batches we can process to meet the order will be N ˆ H K: At each of these N input epochs, we know the following: 1. The outstanding demand to be satis ed, denoted by D, which equals the initial demand minus the total number of good items from all of previous completed and inspected batches. 2. The number of items in each of the K batches still in process, denoted by I 1 ; I 2 ;...; I K. (The subscripts are in reverse order of the sequence by which those batches have been fed into the system; so the next batch whose production will be completed will be of size I K.)
4 374 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371± How many possible input chances are left within the due date; denote this number by n. Then, taking into consideration of the pattern of yield randomness and the relevant costs, we have to choose the number of items for the current input batch (if any). When there are n possible input chances left, we call the corresponding problem an ``n-period decision problem'', or simply an ``n-period problem''. We consider three types of costs: (1) production costs c(x) for a batch of size x (regardless of realized yield), which may include setup costs; (2) an underage (shortage) cost p(á), which applies to the portion of the order not met by the due date and (3) overage costs h(á), which apply to nished good units. h(á) could be negative corresponding to a salvage value. For simplicity, we assume that there are no explicit holding costs for good items produced in early batches. The problem of minimizing the total expected discounted costs can be formulated as a dynamic program. Let f n D; I 1 ; I 2 ;...; I K denote the total expected discounted costs for an n-period problem given the starting state D (the outstanding demand) and I 1 ; I 2 ;...; I K (the sizes of the K batches in process), when an optimal production control rule is used at each of the subsequent decision opportunities. Then, f n D; I 1 ; I 2 ;...; I K satis es the functional equation f n D; I 1 ; I 2 ;...; I K ˆ min a f n 1 D I K u; x; I 1 ;...; I K 1 g u du ; 6 x 6 B c x for n ˆ 1;...; N, where B is the production capacity (if limited) for each batch and a the discount factor per period. Note that the dynamics modeled here is that, when period n ) 1 arrives, every work-in-process batch advances by one step, with I K being completed and x becoming the rst work-in-process batch. Since after the last input batch (i.e., the Nth input epoch) only the outputs from the most recent K batches are still to be realized, the value of f D; I 1 ; I 2 ;...; I K can be calculated by the following recursion: f D; I 1 ; I 2 ;...; I K ˆa f 1 D I K u; ; I 1 ;...; I K 1 g u du; and f i D; ;...; ; I i 1 ;...; I K ˆa f K 1 D; ;...; ; I K ˆ f i 1 D I K u; ;...; ; I i 1 ;...; I K 1 g u du for i ˆ 1; 2;...; K 2 D=I K h I K u D g u du Z D=IK p D I K u g u du: While formulating this very general model is conceptually useful, it turns out to be too complex to analyze. Instead, in the following we analyze a simple version of it, assuming that: 1. There is only one work-in-process batch at any decision making epoch (i.e., K ˆ 1), i.e., the production lead time equals one review period. Denote the size of this batch by I. 2. All cost functions are linear: c ± unit production cost (in particular, setup costs are assumed to equal zero); h ± unit overage cost; p ± unit shortage cost. These are common assumptions in random yield models. 3. There is no production capacity limit. (Note: most of the ``reject allowance'' literature also made this assumption). Since it contains all the essential features of the original problem, we feel that if we can solve this simpli ed problem, we will then gain useful insights into the original complex problem which we have de ned above (it might be only a matter of more complex mathematical manipulations).
5 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371± and The simpli ed problem can be written as follows: f n D; I ˆmin x P cx a f n 1 D Iu; x g u du for n ˆ 1; 2;...; N 1 f D; I ˆh Iu D g u du p D=I Z D=I D Iu g u du: 2 Note D=I can be either less or greater than 1. Regarding the cost parameters, we assume that c < alp (i.e., one would rather input a unit into the process at the last input chance instead of incurring a unit of shortage penalty), and c > alh (that is, one cannot bene t from salvage value alone). As a consequence, we have h p >. De ne V n x; D; I ˆcx a f n 1 D Iu; x g u du for n ˆ 1; 2;...N: 3 Note that, for a given outstanding order D and a work-in-process I, V n (x, D, I) is a function of the current batch size x only. At each decision point, we wish to minimize V n (x, D, I) by choosing the batch size x. Since I is actually the input batch of the previous period, we always have I P. But D may be negative at some later periods, since the total yield (# of good items) from all previous output batches may already exceed the initial order. The reader should note the following notation that will be used in our analysis: of n ; =od and of n ; = always indicate the partial derivatives of f n ; with respect to its rst and second arguments, respectively. The following properties regarding f D; I will be of use: Proposition 1. (i) f D; I is convex in D; I. (ii) o2 f D; I od ˆ o2 f D; I od < : (iii) If D < ; then of D; I od (iv) If D > I; then of D; I od ˆ h and of D; I ˆ p and of D; I ˆ lh: ˆ lp: Proof. See Appendix A. 3. Analysis of one-period problem In this section we provide a complete analysis for a one-period (i.e., only one decision epoch left) decision problem. We show that the optimal policy for choosing the batch size x is a single critical level type with respect to the work-in-process batch size I. Note that this one-period problem looks similar to the classic Newsvendor problem, with the work-in-process batch here roughly corresponding to the outstanding replenishment order in the Newsvendor problem. The key di erence is that the yield of the work-in-process batch (as well as that of the current input batch) here is random. As a consequence, the optimal policy is not
6 376 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371±385 an order-up-to type as in the Newsvendor problem. The yield randomness of the work-in-process batch also di erentiates our problem here from the previous random yield literature (e.g., Henig and Gerchak, 199; Lee and Yano, 1988; Zipkin, 1994, Section 8.4.8), where the yields of all earlier batches are assumed to be known at the current decision epoch. Now, in choosing the batch size x for a one-period problem, the goal, from (3), is to minimize the function V 1 x; D; I ˆcx a We have the following property: f D Iu; x g u du: Proposition 2. V 1 x; D; I is convex in x; D; I. Proof. We know, from Part (i) of Proposition 1, that f D; I is convex in (D, I). The arguments of f D Il; x are linear in (x, D, I). By Theorem 5.7 of Rockafellar (197), f D Il; x is thus convex in (x, D, I). Now, since V 1 x; D; I is a linear combination of convex functions, it is itself convex. Furthermore, it follows from (4) that Z ov 1 x; D; I 1 ˆ c a of D Iu; x =Šg u du 5 ox and Z o 2 V 1 x; D; I 1 ˆ a o 2 f D Iu; x =odšug u du > ; 6 ox where, the inequality follows from Part (ii) of Proposition 1. Eq. (6), in conjunction with Proposition 2, implies that the derivative of V 1 x; D; I with respect to x, i.e., ov 1 x; D; I =ox, is increasing in (x, I). The above simple properties of the function V 1 x; D; I allow us to establish the single critical level nature of optimal policy as follows. Let I 1 D solve ov 1 ; D; I =ox ˆ. That is, I 1 D is given, from (5), by c a of D I 1 D u; =Šg u du ˆ : Since V 1 x; D; I is convex and ov 1 x; D; I =ox is increasing in (x, I), it follows that, in order to minimize V 1 x; D; I, the optimal batch size is x 1 ˆ ifipi 1 D, and it is x 1 > ifi < I 1 D. That is, when the outstanding order is D, I 1 D is the singe critical level for the work-in-process batch size I, which determines whether the current batch size x should be zero or not. When the work-in-process batch size is below its critical level, i.e., I < I 1 D, due to the convexity of V 1 x; D; I, the corresponding optimal batch size x 1 x 1 D; I > is given by equating (5) to zero (the rst order condition). That is c a fof D Iu; x 1 D; I Š=gg u du ˆ : In summary, the optimal decision rule for a one-period lot-sizing problem is as follows: x 1 ˆ x1 D; I ; if I < I 1 D ; ; if I P I 1 D : I 1 D and x 1 D; I solve Eqs. (7) and (8), respectively
7 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371± The following Proposition characterizes the directions in which the critical level I 1 D varies with the outstanding order D, and the optimal batch size x 1 D; I with the outstanding order D and the work-inprocess batch size I. These properties are rather intuitive. Proposition 3. (i) I 1 D is increasing in D. (ii) x 1 D; I is increasing in D and decreasing in I. Proof. See Appendix B. Corresponding to the optimal policy in (9), the value function (the total expected discounted costs) of the one-period problem is given by f 1 D; I ˆmin V 1 x; D; I ˆ V1 x 1 D; I ; D; IŠ if I < I 1 D ; 1 x P V 1 ; D; I if I P I 1 D : We conclude our analysis of the one-period problem by proving the following properties regarding f 1 D; I, which will be useful in our analysis of a two-period problem in Section 4. Proposition 4. i f 1 D; I is convex in D; I : ii of 1 D; I ˆ alh for D < : 11 Proof. Part (i): Since V 1 x; D; I is convex in (x, D, I) (Proposition 2), the convexity of f 1 D; I in (D, I) follows from Theorem 5.7 of Rockafellar (197). Part (ii): When the outstanding order is negative (i.e., D < ), the optimal batch size is obviously zero (i.e., x 1 ˆ ). As a consequence, we have f 1 D; I ˆV 1 ; D; I. From Eq. (4) and part (iii) of Proposition 1, we can then obtain the required result. Note that the second property is intuitive. 4. Analysis of two-period problem In this section, we rst show that the structural properties obtained for the one-period problem also hold for a two-period problem. We then derive some monotonicity properties regarding the critical levels for the work-in-process batch sizes as the number of review periods (i.e., input opportunities) increases. The analysis of a two-period problem will carry over to an n-period problem. From (3), the objective function for a two-period problem is given by V 2 x; D; I ˆcx a f 1 D Iu; x g u du: 12
8 378 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371±385 Proposition 5. V 2 x; D; I is convex in x; D; I. Proof. Using part (i) of Proposition 4, the proof will be similar to that of Proposition 2. To show that a ``single-critical-level'' type policy is optimal for the two-period problem, we will need the following assumption concerning the dependence of f 1 D; I on (D, I): o 2 f 1 D; I od ˆ o2 f 1 D; I od < : Although we proved a similar property for f D; I, as given in Part (ii) of Proposition 1, it turns out to be too complex to prove the inequality part of (13) for f 1 D; I directly in the case I < I 1 D. Nevertheless, we believe intuitively that this condition should hold: While D is the demand (withdrawal) from the system, I tends to increase the supply; hence, the ``e ects'' of D and I to the system should be in opposite directions, as expressed in (13). From (12), we have 13 and ov 2 x; D; I ox ˆ c a of 1 D Iu; x =Šg u du 14 o 2 V 2 x; D; I ox ˆ a o 2 f 1 D Iu; x =odšug u du > ; 15 where, the inequality follows from (13). Thus, ov 2 x; D; I =ox is increasing in (x, I). With the above properties, we can establish the optimality of a single critical level type policy for a twoperiod problem. Let I 2 D solve ov 2 ; D; I =ox ˆ, i.e., c a of 1 D I 2 D u; =Šg u du ˆ : 16 Then, since ov 2 x; D; I =ox is increasing in (x, I), ov 2 ; D; I =ox > for I > I 2 D. This, in conjunction with the convexity of V 2 x; D; I, implies that the optimal batch size for the two-period problem is zero, i.e., x 2 ˆ if the work-in-process batch size is greater than or equal to the critical level; otherwise, x 2 >. When I < I 2 D, due to the convexity of V 2 x; D; I, the corresponding optimal input batch size x 2 x 2 D; I > is obtained by equating (14) to zero, i.e., c a fof 1 D Iu; x 2 D; I Š=gg u du ˆ : In conclusion, we have shown that the optimal decision rule for a two-period lot-sizing problem is again of a ``single-critical-level'' type: x 2 ˆ x2 D; I if I < I 2 D ; 18 if I P I 2 D : One can easily show that I 2 D is increasing in D and that x 2 D; I is increasing in D and decreasing in I. These properties, which are similar to those in Proposition 3 for a one-period problem, are again rather intuitive. The two-step value function is 17
9 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371± f 2 D; I ˆmin 2 x; D; I ˆ V2 x 2 D; I ; D; IŠ if I < I 2 D ; x P V 2 ; D; I if I P I 2 D : 19 Proposition 6. f 2 D; I is convex in D; I. Proof. Using Proposition 5, the proof will be similar to that of Proposition 4. Finally, we derive and discuss some important properties regarding the critical levels for problems with di erent review periods. These properties are summarized in the next Proposition: Proposition 7. (i) I 1 D > I 2 D for all < a 6 1. (ii) I n D > D for a ˆ 1; n ˆ 1; 2. Proof. See Appendix C. We note that the analysis of the two-period problem can actually be extended to an n-period problem. As a consequence, the above Proposition can be extended to (i) I 1 D > I 2 D > > I N D for < a 6 1 and (ii) I n D > D for a ˆ 1; n ˆ 1; 2;...; N, respectively. A possible intuitive explanation for the second result might be as follows. Since the realized yield of any batch will be less than the batch size with probability of one, at any input chance, as long as the size of the work-in-process batch is less than or equal to the outstanding demand D, one should feed some items in the current input batch. But, when the value of a becomes su ciently small, it may then not be worthwhile to input anything, especially at the earlier input chances, due to the discount e ect; so I n D > D is not necessarily true for any value of a. But it turns out to be true for a ˆ 1. The intuition for the property I 1 D > I 2 D > > I N D is as follows. First, when more future input chances are available, one would rather input relatively fewer items and wait to see if the yield of the work-in-process items could actually satisfy D, and thus reduce the risk of larger overage costs; we call this the ``information e ect'' ± information about earlier input batches' yields could be of assistance to the future input decision. Second, when there are more input chances left, the risk of overage and underage could be diversi ed among (or shared by) more input batches; we call this the ``risk diversi cation e ect''. This also explains why it is worthwhile to start production earlier than needed to meet the due date. The bene t of such ``diversi cation over time'' when yields are random was also observed in the context of periodic review models by Henig and Gerchak (199, Theorem 7 and subsequent discussion); see also Zipkin (1994, Section 8.4.8). 5. A numerical example We provide an example to illustrate how the critical levels at di erent review periods are computed. To that end, we assume that the random yield factor U takes the uniform distribution on ; 1Š. That is, g u ˆ1 for 6 u 6 1andl ˆ 1=2. A simple closed-form formula for computing the critical level of an one-period problem, namely, I 1 D, can be derived from Eq. (C.1) in Appendix C as I 1 D ˆa h p 2c ah D: 2
10 38 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371±385 From (C.8), we obtain the constant M as s 2c ah M ˆ : a h p Substituting M into (C.2), we obtain a closed-form formula for computing the critical level of a two-period problem as s I 2 D ˆa2 h p 2c a 2 h 2c ah D: 21 a h p For some concrete results and comparison, we let a ˆ :8; c ˆ 1; h ˆ 5; p ˆ 3 and D ˆ 1: Then, we have I 1 D ˆ117 and I 2 D ˆ86, respectively. If we let a ˆ 1 and keep all the other parameters unchanged, we then have I 1 D ˆ14 and I 2 D ˆ118. Note that I 1 D > I 2 D in both cases and that I 2 D ˆ118 > D ˆ 1 when a ˆ Concluding remarks and future research The single-critical-level (reorder point) but non-order-up-to type policy shown to be optimal in our model is similar to that of the periodic review models with stochastically proportional random yield (e.g., Henig and Gerchak, 199). More interesting properties of the optimal policy obtained here are that I 1 D > I 2 D > > I N D, and that, in the absence of discounting, I n D > D. I n D > D implies that the total input quantity (work-in-process I plus the input x for the current batch) will be greater than the outstanding demand D. Most of the reject allowance literature have assumed that the input lot is always greater than the outstanding demand. While this seems rather intuitive and was proven to be true in the case of Binomial yield by Beja (1977), Grosfeld-Nir and Gerchak (1996) provide examples which show that it is not true in general. Several possible extensions to our current model are of interest both theoretically and practically. Like most of the reject allowance literature, we have assumed that the demand is deterministic. This may not always be the case in practice. We believe that a single critical level type policy will still be optimal when the demand is random. To incorporate setup costs into the current model and provide an analytical solution would be a di cult problem. This being so, some heuristics may need to be devised. Appendix A. Proof of Proposition 1 From Eq. (2), we have, after some algebra, that of D; I od ˆ h h p G D=I ; A:1 o 2 f D; I od 2 ˆ h p g D=I =I > ; A:2
11 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371± Z of D; I D=I ˆ lh h p ug u du; A:3 o 2 f D; I 2 ˆ h p D 2 g D=I =I 3 > ; A:4 o 2 f D; I od ˆ o2 f D; I od ˆ h p Dg D=I =I 2 < ; A:5 o 2 f D; I od 2 o2 f D; I o2 f D; I 2 od o2 f D; I od ˆ : A:6 Thus, Part (i) of the Proposition follows from (A.2), (A.4) and (A.6), and Part (ii) follows from (A.4). Parts (iii) and (iv) will follow from a direct check of (A.1) and (A.3). Appendix B. Proof of Proposition 3 Part (i): From (7), it follows, after some algebra, that di 1 D dd ˆ R 1 fo2 f D I 1 D u; Š=oDgg u du R 1 fo2 f D I 1 D u; Š=oDgug u du > ; where the inequality follows from Part (ii) of Proposition 1. Part (ii): From (8), we have that and ox 1 D; I od ox 1 D; I ˆ ˆ R 1 fo2 f D I 1 D u; Š=oDgg u du R 1 fo2 f D I 1 D u; Š= 2 gg u du > ; R 1 fo2 f D I 1 D u; Š=oDgug u du R 1 fo2 f D I 1 D u; Š= 2 gg u du < ; where the inequalities follow from Parts (ii) and (i) of Proposition 1 (which implies that o 2 f D I 1 D u; Š= 2 > ). Hence, x 1 D; I is increasing in D and decreasing in I. Appendix C. Proof of Proposition 7 Part (i): By Parts (iii) and (iv) of Proposition 1, the left-hand side of (7) can be written as
12 382 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371±385 Z D=I1 D ov 1 ; D; I 1 D; I Š ˆ c a fof D I 1 D u; Š=gg u du ox a fof D I 1 D u; Š=gg u du D=I 1 D ˆ c a Z D=I1 D So, I 1 D is determined by the equation lp g u du a ˆ c alf h p G D=I 1 D Š hg: D=I 1 D lh g u du c alf h p G D=I 1 D Š hg ˆ: C:1 Eq. (16), by which I 2 D is determined, can be written as (detailed proof follows) c a 2 lf h p G M G D=I 2 D Š hg ˆ; C:2 where, M is a given constant with < M < 1. Comparing (C.1) with (C.2), it is clear that each of the two extra factors a and G M in (C.2), both of which are between and 1, will force I 2 D to become less than I 1 D. That is, in general we will have I 1 D > I 2 D. Now, we show the equivalence of (16) and (C.2). That is, we need to prove that ov 2 ; D; I 2 D Š ˆ c a 2 lf h p G M G D=I 2 D Š hg: ox First, when D > and I ˆ, we have x 1 D; I ˆx 1 D; >. From (8), x 1 D; solves that is, c a fof D; x 1 D; Š=gg u du ˆ ; c afof D; x 1 D; Š=g ˆ: From (2), we have of D; I ˆ lh h p Z D=I of D; I ˆ h h p G D=I : od Substituting (C.5) into (C.4), we get c alh a h p Z D=x1 D; ug u du; ug u du ˆ : De ne D=x 1 D; M in (C.7) for all D >. So, M is a constant determined by c alh a h p ug u du ˆ : We can easily verify that < M < 1, in order to satisfy (C.8). C:3 C:4 C:5 C:6 C:7 C:8
13 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371± Second, for I < I 1 D, we have from (1) that f 1 D; I ˆV 1 x 1 D; I ; D; IŠ. It follows that of 1 D; I ˆ ov 1 x 1 D; I ; D; IŠ ˆ ov 1 x 1 D; I ; D; IŠ ˆ a ov 1 x 1 D; I ; D; IŠ ox 1 D; I ox fof D Iu; x 1 D; I Š=oDgug u du; where, the second equality follows from the optimality of x 1 D; I, and the third from (4). When D > and I ˆ, this reduces to of 1 D; ˆ a fof D; x 1 D; Š=oDgug u du ˆ alfof D; x 1 D; Š=oDg ˆ alf h h p G D=x 1 D; Šg ˆ al h h p G M Š; C:9 where, the third equality follows from (C.6) and M is de ned in (C.8). Third, using (C.9) and (11) for of 1 D; = when D > and D <, respectively, we can write, from (14), that ov 2 ; D; I 2 D Š ox ˆ c a ˆ c a a ˆ c a a fof 1 D I 2 D u; Š=gg u du Z D=I2 D Z D=I2 D Z D=I2 D Z D=I2 D fof 1 D I 2 D u; Š=gg u du fof 1 D I 2 D u; Š=gg u du f al h h p G M Šgg u du alh g u du ˆ c a 2 lf h p G M G D=I 2 D Š hg; which gives us (C.3). We thus completed the proof of Part (i). Part (ii): For a ˆ 1, we rst show that I 1 D > D and then I 2 D > D. Assume I 1 D < D. With a ˆ 1, the left-hand side of (C.1) would reduce to c lp < ˆ right-hand side (where the inequality follows by our assumption c alp < about the values of the cost parameters). Hence, we must have I 1 D > D in order to satisfy (C.1), since ov 1 x; D; I =ox is increasing in I. When a ˆ 1, the next argument shows that the left-hand side of (C.2) becomes negative if I 2 D < D, and as a consequence, we must also have I 2 D > D. IfI 2 D < D, simply substituting a ˆ 1 and G D=I 2 D Š ˆ 1 into (C.2), we have
14 384 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371±385 Left-hand ˆ c lh l h p G M ˆ h p ug u du l h p G M ˆ h p ug u du l g u du < ; where the second equality is due to c lh ˆ h p R M ug u du from (C.8); and the inequality follows from Lemma 1 (see below). We thus completed the proof. Lemma 1. If U is a random variable on [, 1] with a density g(á) and mean l, then for any given value of M, <M <1, ug u du l g u du < : Proof. Consider the following two cases: (a) For < M < l, ug u du l g u du < l ug u du l g u du ˆ : (b) For l < M < 1, let F M ˆ ug u du l g u du: It follows that F M ˆMg M lg M ˆ M l g M > ; that is, F(M) is increasing in M. In conjunction with the fact of F(1) ˆ, we have F M < for l < M < 1. References Barad, M., Braha, D., Control limits for multi-stage manufacturing process with binomial yield single and multiple production runs. Journal of the Operational Research Society 47, 98±112. Beja, A., Optimal reject allowance with constant marginal production e ciency. Naval Research Logistics Quarterly 24, 21±33. Chu, C., Proth, J.M., Xie, X., Supply chain management in assembly systems. Naval Research Logistics 4, 933±95. Gerchak, Y., Wang, Y., Yano, C.A., Lot sizing in assembly systems with random component yields. IIE Transactions 26, 19±24. Grosfeld-Nir, A., Gerchak, Y., 199. Multiple lot sizing with random common-cause yield and rigid demand. Operations Research Letters 9, 383±388. Grosfeld-Nir, A., Gerchak, Y., Production to order with random yields: single-stage multiple lot sizing. IIE Transactions 28, 669±676. Grosfeld-Nir, A., Gerchak, Y., Multi-Stage Production to Order with Rework Capability, Working Paper, Department of Management Sciences, University of Waterloo, Ontario, Canada. Grosfeld-Nir, A., Ronen, B., A single bottleneck system with binomial yield and rigid demand. Management Science 39, 65± 653. Gurnani, H., Akella, R., Lehoczky, J., Optimal order policies in assembly systems with random demand and random supplier delivery. IIE Transactions 28, 865±878.
15 Y. Wang, Y. Gerchak / European Journal of Operational Research 126 (2) 371± Henig, M., Gerchak, Y., 199. The structure of periodic review policies in the presence of variable yield. Operations Research 38, 634± 643. Klein, M., Markovian decision models for reject allowance problems. Management Science 12, 349±358. Lee, H.L., Yano, C.A., Production control for multi-stage systems with variable yield losses. Operations Research 36, 269±278. Proth, J.M., Mauroy, G., Wardi, Y., Chu, C., Xie, X.L., Supply management for cost minimization in assembly systems with random component yield times. Journal of Intelligent Manufacturing 8, 385±43. Rockafellar R.T., 197. Convex Analysis. Princeton University Press, Princeton, NJ. Singh, M.R., Abraham, C.T., Akella, R., 199. A wafer design problem in semiconductor manufacturing for reliable customer service. IEEE Transactions on CHMT 13, 13±18. Sepehri, M., Silver, E.A., New, C., A heuristic for multiple lot sizing for an order under variable yield. IIE Transactions 18, 63± 69. Wang, Y., Gerchak, Y., Periodic review production models with variable capacity, random yield, and uncertain demand. Management Science 42, 13±137. Wein, A.S., Random yield, rework and scrap in multi-stage batch manufacturing environments. Operations Research 4, 551± 563. White, D.J., Dynamic programming and systems of uncertain duration. Management Science 12, 37±67. Yano, C.A., Lee, H.L. Lot sizing with random yields a review. Operations Research 43, 311±334. Zipkin, P.H., Foundations of inventory management. Unpublished Manuscript.
The multi-item single period problem with an initial stock of convertible units
European Journal of Operational Research 132 2001) 466±477 www.elsevier.com/locate/dsw Theory and Methodology The multi-item single period problem with an initial stock of convertible units Edward A. Silver
More informationProduction Policies for Multi-Product Systems with Deteriorating. Process Condition
Production Policies for Multi-Product Systems with Deteriorating Process Condition Burak Kazaz School of Business, University of Miami, Coral Gables, FL 3324. bkazaz@miami.edu Thomas W. Sloan College of
More informationAn optimal batch size for a production system under linearly increasing time-varying demand process
Computers & Industrial Engineering 4 (00) 35±4 www.elsevier.com/locate/dsw An optimal batch size for a production system under linearly increasing time-varying demand process Mohd Omar a, *, David K. Smith
More informationIntegrating Quality and Inspection for the Optimal Lot-sizing Problem with Rework, Minimal Repair and Inspection Time
Proceedings of the 2011 International Conference on Industrial Engineering and Operations Management Kuala Lumpur, Malaysia, January 22 24, 2011 Integrating Quality and Inspection for the Optimal Lot-sizing
More informationMinimum Wages and Excessive E ort Supply
Minimum Wages and Excessive E ort Supply Matthias Kräkel y Anja Schöttner z Abstract It is well-known that, in static models, minimum wages generate positive worker rents and, consequently, ine ciently
More informationEffect of repair cost on the rework decision-making in an economic production quantity model
Effect of repair cost on the rework decision-making in an economic production quantity model Singa Wang Chiu Department of Business Administration Chaoyang University of Technology 68, Gifeng E. Rd. Wufeng,
More informationLabor Economics, Lecture 11: Partial Equilibrium Sequential Search
Labor Economics, 14.661. Lecture 11: Partial Equilibrium Sequential Search Daron Acemoglu MIT December 6, 2011. Daron Acemoglu (MIT) Sequential Search December 6, 2011. 1 / 43 Introduction Introduction
More informationTHIELE CENTRE. The M/M/1 queue with inventory, lost sale and general lead times. Mohammad Saffari, Søren Asmussen and Rasoul Haji
THIELE CENTRE for applied mathematics in natural science The M/M/1 queue with inventory, lost sale and general lead times Mohammad Saffari, Søren Asmussen and Rasoul Haji Research Report No. 11 September
More informationA Hierarchy of Suboptimal Policies for the Multi-period, Multi-echelon, Robust Inventory Problem
A Hierarchy of Suboptimal Policies for the Multi-period, Multi-echelon, Robust Inventory Problem Dimitris J. Bertsimas Dan A. Iancu Pablo A. Parrilo Sloan School of Management and Operations Research Center,
More informationOptimal Acquisition Quantities in Remanufacturing with Condition Uncertainty
Optimal Acquisition Quantities in Remanufacturing with Condition Uncertainty Michael R. Galbreth Moore School of Business University of South Carolina 1705 College Street Columbia, SC, 29208 Tel: (803)777-4242
More informationSupply planning optimization for linear production system with stochastic lead-times and quality control
Supply planning optimization for linear production system with stochastic lead-times and quality control O Ben Ammar, B Bettayeb, Alexandre Dolgui To cite this version: O Ben Ammar, B Bettayeb, Alexandre
More informationScheduling jobs with agreeable processing times and due dates on a single batch processing machine
Theoretical Computer Science 374 007 159 169 www.elsevier.com/locate/tcs Scheduling jobs with agreeable processing times and due dates on a single batch processing machine L.L. Liu, C.T. Ng, T.C.E. Cheng
More informationCoordinated Replenishments at a Single Stocking Point
Chapter 11 Coordinated Replenishments at a Single Stocking Point 11.1 Advantages and Disadvantages of Coordination Advantages of Coordination 1. Savings on unit purchase costs.. Savings on unit transportation
More informationThe Kuhn-Tucker Problem
Natalia Lazzati Mathematics for Economics (Part I) Note 8: Nonlinear Programming - The Kuhn-Tucker Problem Note 8 is based on de la Fuente (2000, Ch. 7) and Simon and Blume (1994, Ch. 18 and 19). The Kuhn-Tucker
More informationResearch Article A Partial Backlogging Inventory Model for Deteriorating Items with Fluctuating Selling Price and Purchasing Cost
Advances in Operations Research Volume 2012, Article ID 385371, 15 pages doi:10.1155/2012/385371 Research Article A Partial Backlogging Inventory Model for Deteriorating Items with Fluctuating Selling
More informationECON0702: Mathematical Methods in Economics
ECON0702: Mathematical Methods in Economics Yulei Luo SEF of HKU January 14, 2009 Luo, Y. (SEF of HKU) MME January 14, 2009 1 / 44 Comparative Statics and The Concept of Derivative Comparative Statics
More informationLecture 6: Contraction mapping, inverse and implicit function theorems
Lecture 6: Contraction mapping, inverse and implicit function theorems 1 The contraction mapping theorem De nition 11 Let X be a metric space, with metric d If f : X! X and if there is a number 2 (0; 1)
More informationInventory management with advance capacity information
Inventory management with advance capacity information Jakšic, M.; Fransoo, J.C.; Tan, T.; de Kok, A.G.; Rusjan, B. Published: 01/01/2008 Document Version Accepted manuscript including changes made at
More informationAdmission control schemes to provide class-level QoS in multiservice networks q
Computer Networks 35 (2001) 307±326 www.elsevier.com/locate/comnet Admission control schemes to provide class-level QoS in multiservice networks q Suresh Kalyanasundaram a,1, Edwin K.P. Chong b, Ness B.
More informationEconomics of Controlling Climate Change under Uncertainty.
Economics of Controlling Climate Change under Uncertainty. Alexander Golub y Environmental Defense Fund, Washington, DC. Santanu Roy z Southern Methodist University, Dallas, TX. October 18, 2010 Abstract
More informationParallel machine scheduling with batch delivery costs
Int. J. Production Economics 68 (2000) 177}183 Parallel machine scheduling with batch delivery costs Guoqing Wang*, T.C. Edwin Cheng Department of Business Administration, Jinan University, Guangzhou,
More informationECON2285: Mathematical Economics
ECON2285: Mathematical Economics Yulei Luo Economics, HKU September 17, 2018 Luo, Y. (Economics, HKU) ME September 17, 2018 1 / 46 Static Optimization and Extreme Values In this topic, we will study goal
More informationOptimal inventory policies with non-stationary supply disruptions and advance supply information Atasoy, B.; Güllü, R.; Tan, T.
Optimal inventory policies with non-stationary supply disruptions and advance supply information Atasoy, B.; Güllü, R.; Tan, T. Published: 01/01/2011 Document Version Publisher s PDF, also known as Version
More informationMultistage pulse tubes
Cryogenics 40 (2000) 459±464 www.elsevier.com/locate/cryogenics Multistage pulse tubes A.T.A.M. de Waele *, I.A. Tanaeva, Y.L. Ju Department of Physics, Eindhoven University of Technology, P.O. Box 513,
More informationAppendix for "O shoring in a Ricardian World"
Appendix for "O shoring in a Ricardian World" This Appendix presents the proofs of Propositions - 6 and the derivations of the results in Section IV. Proof of Proposition We want to show that Tm L m T
More informationOptimal Control of an Inventory System with Joint Production and Pricing Decisions
Optimal Control of an Inventory System with Joint Production and Pricing Decisions Ping Cao, Jingui Xie Abstract In this study, we consider a stochastic inventory system in which the objective of the manufacturer
More informationExpected Time Delay in Multi-Item Inventory Systems with Correlated Demands
Expected Time Delay in Multi-Item Inventory Systems with Correlated Demands Rachel Q. Zhang Department of Industrial and Operations Engineering, University of Michigan, Ann Arbor, Michigan 48109 Received
More informationAddendum to: International Trade, Technology, and the Skill Premium
Addendum to: International Trade, Technology, and the Skill remium Ariel Burstein UCLA and NBER Jonathan Vogel Columbia and NBER April 22 Abstract In this Addendum we set up a perfectly competitive version
More informationCoordinating Inventory Control and Pricing Strategies with Random Demand and Fixed Ordering Cost: The Finite Horizon Case
OPERATIONS RESEARCH Vol. 52, No. 6, November December 2004, pp. 887 896 issn 0030-364X eissn 1526-5463 04 5206 0887 informs doi 10.1287/opre.1040.0127 2004 INFORMS Coordinating Inventory Control Pricing
More informationLecture Notes - Dynamic Moral Hazard
Lecture Notes - Dynamic Moral Hazard Simon Board and Moritz Meyer-ter-Vehn October 27, 2011 1 Marginal Cost of Providing Utility is Martingale (Rogerson 85) 1.1 Setup Two periods, no discounting Actions
More informationAppendix - E-Companion
Article submitted to Operations Research; manuscript no. Please, provide the manuscript number! 1 Appendix - E-Companion Appendix A: Derivation of optimal supply volume The supply volume x 1 was treated
More informationInventory control with partial batch ordering
Inventory control with partial batch ordering Alp, O.; Huh, W.T.; Tan, T. Published: 01/01/2009 Document Version Publisher s PDF, also known as Version of Record (includes final page, issue and volume
More informationAn Uncertain Bilevel Newsboy Model with a Budget Constraint
Journal of Uncertain Systems Vol.12, No.2, pp.83-9, 218 Online at: www.jus.org.uk An Uncertain Bilevel Newsboy Model with a Budget Constraint Chunliu Zhu, Faquan Qi, Jinwu Gao School of Information, Renmin
More informationStochastic Analysis of Bidding in Sequential Auctions and Related Problems.
s Case Stochastic Analysis of Bidding in Sequential Auctions and Related Problems S keya Rutgers Business School s Case 1 New auction models demand model Integrated auction- inventory model 2 Optimizing
More informationThe increased complexity of modern manufacturing has led to uncertainties in production
Optimal Production Policies for Multi-Stage Systems with Setup Costs and Uncertain Capacities Juhwen Hwang j Medini R. Singh Institute of Business Management, National Dong Hwa University, Hualien, Taiwan,
More informationOn the synchronization of a class of electronic circuits that exhibit chaos
Chaos, Solitons and Fractals 13 2002) 1515±1521 www.elsevier.com/locate/chaos On the synchronization of a class of electronic circuits that exhibit chaos Er-Wei Bai a, *, Karl E. Lonngren a, J.C. Sprott
More informationOPTIMALITY OF RANDOMIZED TRUNK RESERVATION FOR A PROBLEM WITH MULTIPLE CONSTRAINTS
OPTIMALITY OF RANDOMIZED TRUNK RESERVATION FOR A PROBLEM WITH MULTIPLE CONSTRAINTS Xiaofei Fan-Orzechowski Department of Applied Mathematics and Statistics State University of New York at Stony Brook Stony
More informationStochastic (Random) Demand Inventory Models
Stochastic (Random) Demand Inventory Models George Liberopoulos 1 The Newsvendor model Assumptions/notation Single-period horizon Uncertain demand in the period: D (parts) assume continuous random variable
More informationLecture 3 - Axioms of Consumer Preference and the Theory of Choice
Lecture 3 - Axioms of Consumer Preference and the Theory of Choice David Autor 14.03 Fall 2004 Agenda: 1. Consumer preference theory (a) Notion of utility function (b) Axioms of consumer preference (c)
More informationImproving Supply Chain Performance: Real-Time Demand Information and Flexible Deliveries
Improving Supply Chain Performance: Real-Time Demand Information and Flexible Deliveries Kevin H. Shang Sean X. Zhou Geert-Jan van Houtum The Fuqua School of Business, Duke University, Durham, North Carolina
More informationSupplemental Material 1 for On Optimal Inference in the Linear IV Model
Supplemental Material 1 for On Optimal Inference in the Linear IV Model Donald W. K. Andrews Cowles Foundation for Research in Economics Yale University Vadim Marmer Vancouver School of Economics University
More informationNewsvendor Model with Random Supply
Newsvendor Model with Random Suppl H.K. Oka, F. Karaesmen and S. Özekici Koç Universit Department of Industrial Engineering 3445 Sar er-istanbul, Turke September 21 Abstract The newsvendor model is perhaps
More informationSingle-part-type, multiple stage systems
MIT 2.853/2.854 Introduction to Manufacturing Systems Single-part-type, multiple stage systems Stanley B. Gershwin Laboratory for Manufacturing and Productivity Massachusetts Institute of Technology Single-stage,
More informationOrdering Policies for Periodic-Review Inventory Systems with Quantity-Dependent Fixed Costs
OPERATIONS RESEARCH Vol. 60, No. 4, July August 2012, pp. 785 796 ISSN 0030-364X (print) ISSN 1526-5463 (online) http://dx.doi.org/10.1287/opre.1110.1033 2012 INFORMS Ordering Policies for Periodic-Review
More informationMachine Scheduling with Deliveries to Multiple Customer Locations
This is the Pre-Published Version. Machine Scheduling with Deliveries to Multiple Customer Locations Chung-Lun Li George Vairaktarakis Chung-Yee Lee April 2003 Revised: November 2003 Abstract One important
More informationResearch Article Investing in Lead-Time Variability Reduction in a Quality-Adjusted Inventory Model with Finite-Range Stochastic Lead-Time
Hindawi Publishing Corporation Journal of Applied Mathematics and Decision Sciences Volume 2008, Article ID 795869, 13 pages doi:10.1155/2008/795869 Research Article Investing in Lead-Time Variability
More informationBanks, depositors and liquidity shocks: long term vs short term interest rates in a model of adverse selection
Banks, depositors and liquidity shocks: long term vs short term interest rates in a model of adverse selection Geethanjali Selvaretnam Abstract This model takes into consideration the fact that depositors
More informationOptimal Control of Stochastic Inventory System with Multiple Types of Reverse Flows. Xiuli Chao University of Michigan Ann Arbor, MI 48109
Optimal Control of Stochastic Inventory System with Multiple Types of Reverse Flows Xiuli Chao University of Michigan Ann Arbor, MI 4809 NCKU Seminar August 4, 009 Joint work with S. Zhou and Z. Tao /44
More informationA dynamic tra c equilibrium assignment paradox
Transportation Research Part B 34 (2) 515±531 www.elsevier.com/locate/trb A dynamic tra c equilibrium assignment paradox Takashi Akamatsu * Department of Knowledge-based Information Engineering, Toyohashi
More informationSolow Growth Model. Michael Bar. February 28, Introduction Some facts about modern growth Questions... 4
Solow Growth Model Michael Bar February 28, 208 Contents Introduction 2. Some facts about modern growth........................ 3.2 Questions..................................... 4 2 The Solow Model 5
More informationCost models for lot streaming in a multistage flow shop
Omega 33 2005) 435 450 www.elsevier.com/locate/omega Cost models for lot streaming in a multistage flow shop Huan Neng Chiu, Jen Huei Chang Department of Industrial Management, National Taiwan University
More informationStochastic Optimization
Chapter 27 Page 1 Stochastic Optimization Operations research has been particularly successful in two areas of decision analysis: (i) optimization of problems involving many variables when the outcome
More informationLinearized methods for ordinary di erential equations
Applied Mathematics and Computation 104 (1999) 109±19 www.elsevier.nl/locate/amc Linearized methods for ordinary di erential equations J.I. Ramos 1 Departamento de Lenguajes y Ciencias de la Computacion,
More informationAppointment Scheduling with Discrete Random Durations
MATHEMATICS OF OPERATIONS RESEARCH Vol. 36, No. 2, May 2011, pp. 240 257 issn 0364-765X eissn 1526-5471 11 3602 0240 doi 10.1287/moor.1110.0489 2011 INFORMS Appointment Scheduling with Discrete Random
More informationStochastic dominance with imprecise information
Stochastic dominance with imprecise information Ignacio Montes, Enrique Miranda, Susana Montes University of Oviedo, Dep. of Statistics and Operations Research. Abstract Stochastic dominance, which is
More informationHeuristics for the two-stage job shop scheduling problem with a bottleneck machine
European Journal of Operational Research 123 (2000) 229±240 www.elsevier.com/locate/orms Heuristics for the two-stage job shop scheduling problem with a bottleneck machine I.G. Drobouchevitch, V.A. Strusevich
More informationNo-Holdback Allocation Rules for Continuous-Time Assemble-to-Order Systems
OPERATIONS RESEARCH Vol. 58, No. 3, May June 2010, pp. 691 705 issn 0030-364X eissn 1526-5463 10 5803 0691 informs doi 10.1287/opre.1090.0785 2010 INFORMS No-Holdback Allocation Rules for Continuous-Time
More informationASIGNIFICANT research effort has been devoted to the. Optimal State Estimation for Stochastic Systems: An Information Theoretic Approach
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, VOL 42, NO 6, JUNE 1997 771 Optimal State Estimation for Stochastic Systems: An Information Theoretic Approach Xiangbo Feng, Kenneth A Loparo, Senior Member, IEEE,
More informationOptimal taxation with monopolistic competition
Optimal taxation with monopolistic competition Leslie J. Reinhorn Economics Department University of Durham 23-26 Old Elvet Durham DH1 3HY United Kingdom phone +44 191 334 6365 fax +44 191 334 6341 reinhorn@hotmail.com
More informationCoordinated Quality Control in a Two-Stage System
This is the Pre-Published Version Coordinated Quality Control in a Two-Stage System David D. Yao Department of Industrial Engineering and Operations Research Columbia University, New York, NY 10027-6699
More informationAn optimal policy for a single-vendor single-buyer integrated production}inventory system with capacity constraint of the transport equipment
Int. J. Production Economics 65 (2000) 305}315 Technical Note An optimal policy f a single-vend single-buyer integrated production}inventy system with capacity constraint of the transpt equipment M.A.
More informationNew Markov Decision Process Formulations and Optimal Policy Structure for Assemble-to-Order and New Product Development Problems
New Markov Decision Process Formulations and Optimal Policy Structure for Assemble-to-Order and New Product Development Problems by EMRE NADAR Submitted to the Tepper School of Business in Partial Fulfillment
More informationReplenishment Planning for Stochastic Inventory System with Shortage Cost
Replenishment Planning for Stochastic Inventory System with Shortage Cost Roberto Rossi UCC, Ireland S. Armagan Tarim HU, Turkey Brahim Hnich IUE, Turkey Steven Prestwich UCC, Ireland Inventory Control
More informationInfluence of product return lead-time on inventory control
Influence of product return lead-time on inventory control Mohamed Hichem Zerhouni, Jean-Philippe Gayon, Yannick Frein To cite this version: Mohamed Hichem Zerhouni, Jean-Philippe Gayon, Yannick Frein.
More informationTime. 3 p3 Time. (a) (b)
A Hybrid Algorithm for Solving the Economic Lot and Delivery Scheduling Problem in the Common Cycle Case Suquan Ju Jens Clausen Informatics and Mathematical Modelling echnical University of Denmark 2800
More informationExperimentation and Observational Learning in a Market with Exit
ömmföäflsäafaäsflassflassflas ffffffffffffffffffffffffffffffffffff Discussion Papers Experimentation and Observational Learning in a Market with Exit Pauli Murto Helsinki School of Economics and HECER
More informationMulti-stage newsboy problem: A dynamic model
European Journal of Operational Research 49 (23) 448 458 Production, Manufacturing and Logistics Multi-stage newsboy problem: A dynamic model Konstantin Kogan a, *, Sheldon Lou b a Department of Interdisciplinary
More information2 Interval-valued Probability Measures
Interval-Valued Probability Measures K. David Jamison 1, Weldon A. Lodwick 2 1. Watson Wyatt & Company, 950 17th Street,Suite 1400, Denver, CO 80202, U.S.A 2. Department of Mathematics, Campus Box 170,
More informationThe Single and Multi-Item Transshipment Problem with Fixed Transshipment Costs
The Single and Multi-Item Transshipment Problem with Fixed Transshipment Costs Reut Noham, Michal Tzur Department of Industrial Engineering, Tel-Aviv University, Tel Aviv, Israel Received 1 October 2013;
More informationSolving Extensive Form Games
Chapter 8 Solving Extensive Form Games 8.1 The Extensive Form of a Game The extensive form of a game contains the following information: (1) the set of players (2) the order of moves (that is, who moves
More informationDynamic Markets for Lemons: Performance, Liquidity, and Policy Intervention
Dynamic Markets for Lemons: Performance, Liquidity, and Policy Intervention Diego Moreno y John Wooders z August 2013 Abstract We study non-stationary dynamic decentralized markets with adverse selection
More informationAsymptotically Optimal Inventory Control For Assemble-to-Order Systems
Asymptotically Optimal Inventory Control For Assemble-to-Order Systems Marty Reiman Columbia Univerisity joint work with Mustafa Dogru, Haohua Wan, and Qiong Wang May 16, 2018 Outline The Assemble-to-Order
More informationSimple Policies for Joint Replenishment Can Perform Badly
Simple Policies for Joint Replenishment Can Perform Badly Christopher R. Dance, Onno R. Zoeter, Haengju Lee Abstract We consider the stochastic joint replenishment problem in which several items must be
More informationFurther discussion on linear production functions and DEA
European Journal of Operational Research 127 (2000) 611±618 www.elsevier.com/locate/dsw Theory and Methodology Further discussion on linear production functions and DEA Joe Zhu * Department of Management,
More informationF E M M Faculty of Economics and Management Magdeburg
OTTO-VON-GUERICKE-UNIVERSITY MAGDEBURG FACULTY OF ECONOMICS AND MANAGEMENT On the alignment of lot sizing decisions in a remanufacturing system in the presence of random yield Tobias Schulz Ivan Ferretti
More informationLecture Notes - Dynamic Moral Hazard
Lecture Notes - Dynamic Moral Hazard Simon Board and Moritz Meyer-ter-Vehn October 23, 2012 1 Dynamic Moral Hazard E ects Consumption smoothing Statistical inference More strategies Renegotiation Non-separable
More informationEconomics 620, Lecture 18: Nonlinear Models
Economics 620, Lecture 18: Nonlinear Models Nicholas M. Kiefer Cornell University Professor N. M. Kiefer (Cornell University) Lecture 18: Nonlinear Models 1 / 18 The basic point is that smooth nonlinear
More informationThe Intuitive and Divinity Criterion:
The Intuitive and Divinity Criterion: Interpretation and Step-by-Step Examples Ana Espínola-Arredondo School of Economic Sciences Washington State University Pullman, WA 99164 Félix Muñoz-García y School
More informationInventory systems with variable capacity. Ilkyeong Moon and Byung-Hyun Ha*
68 European J. Industrial Engineering, Vol. 6, No. 1, 212 Inventory systems with variable capacity Ilkyeong Moon and Byung-Hyun Ha* Department of Industrial Engineering, Pusan National University, Busan,
More informationLecture Notes 8
14.451 Lecture Notes 8 Guido Lorenzoni Fall 29 1 Stochastic dynamic programming: an example We no turn to analyze problems ith uncertainty, in discrete time. We begin ith an example that illustrates the
More informationc 2007 Je rey A. Miron
Review of Calculus Tools. c 007 Je rey A. Miron Outline 1. Derivatives. Optimization 3. Partial Derivatives. Optimization again 5. Optimization subject to constraints 1 Derivatives The basic tool we need
More informationResearch Article Batch Scheduling on Two-Machine Flowshop with Machine-Dependent Setup Times
Advances in Operations Research Volume 2009, Article ID 153910, 10 pages doi:10.1155/2009/153910 Research Article Batch Scheduling on Two-Machine Flowshop with Machine-Dependent Setup Times Lika Ben-Dati,
More informationA Duality-Based Relaxation and Decomposition Approach for Inventory Distribution Systems
A Duality-Based Relaxation and Decomposition Approach for Inventory Distribution Systems Sumit Kunnumkal, 1 Huseyin Topaloglu 2 1 Indian School of Business, Gachibowli, Hyderabad 500032, India 2 School
More informationGMM-based inference in the AR(1) panel data model for parameter values where local identi cation fails
GMM-based inference in the AR() panel data model for parameter values where local identi cation fails Edith Madsen entre for Applied Microeconometrics (AM) Department of Economics, University of openhagen,
More informationOptimal Policies for an Assemble-to-Order N-System
Optimal Policies for an Assemble-to-Order N-System Lijian Lu Jing-Sheng Song Hanqin Zhang NUS Business School, National University of Singapore, Singapore Fuqua School of Business, Duke University, Durham,
More informationStrategy-proof allocation of indivisible goods
Soc Choice Welfare (1999) 16: 557±567 Strategy-proof allocation of indivisible goods Lars-Gunnar Svensson Department of Economics, Lund University, P.O. Box 7082, SE-220 07 of Lund, Sweden (e-mail: lars-gunnar.svensson@nek.lu.se)
More informationA Proof of the EOQ Formula Using Quasi-Variational. Inequalities. March 19, Abstract
A Proof of the EOQ Formula Using Quasi-Variational Inequalities Dir Beyer y and Suresh P. Sethi z March, 8 Abstract In this paper, we use quasi-variational inequalities to provide a rigorous proof of the
More informationA note on Two-warehouse inventory model with deterioration under FIFO dispatch policy
Available online at www.sciencedirect.com European Journal of Operational Research 190 (2008) 571 577 Short Communication A note on Two-warehouse inventory model with deterioration under FIFO dispatch
More informationProofs for Stress and Coping - An Economic Approach Klaus Wälde 56 October 2017
A Appendix Proofs for Stress and Coping - An Economic Approach Klaus älde 56 October 2017 A.1 Solution of the maximization problem A.1.1 The Bellman equation e start from the general speci cation of a
More informationMACHINE DEDICATION UNDER PRODUCT AND PROCESS DIVERSITY. Darius Rohan. IBM Microelectonics Division East Fishkill, NY 12533, U.S.A.
Proceedings of the 1999 Winter Simulation Conference P. A. Farrington, H. B. Nembhard, D. T. Sturrock, and G. W. Evans, eds. MACHINE DEDICATION UNDER PRODUCT AND PROCESS DIVERSITY Darius Rohan IBM Microelectonics
More information7.1 INTRODUCTION. In this era of extreme competition, each subsystem in different
7.1 INTRODUCTION In this era of extreme competition, each subsystem in different echelons of integrated model thrives to improve their operations, reduce costs and increase profitability. Currently, the
More informationAn alternative theorem for generalized variational inequalities and solvability of nonlinear quasi-p M -complementarity problems
Applied Mathematics and Computation 109 (2000) 167±182 www.elsevier.nl/locate/amc An alternative theorem for generalized variational inequalities and solvability of nonlinear quasi-p M -complementarity
More informationThird down with a yard to go : the Dixit-Skeath conundrum on equilibria in competitive games.
June 28, 1999 Third down with a yard to go : the Dixit-Skeath conundrum on equilibria in competitive games. Abstract In strictly competitive games, equilibrium mixed strategies are invariant to changes
More information2001, Dennis Bricker Dept of Industrial Engineering The University of Iowa. DP: Producing 2 items page 1
Consider a production facility which can be devoted in each period to one of two products. For simplicity, we assume that the production rate is deterministic and that production is always at full capacity.
More informationOne important issue in the study of queueing systems is to characterize departure processes. Study on departure processes was rst initiated by Burke (
The Departure Process of the GI/G/ Queue and Its MacLaurin Series Jian-Qiang Hu Department of Manufacturing Engineering Boston University 5 St. Mary's Street Brookline, MA 2446 Email: hqiang@bu.edu June
More informationExclusive contracts and market dominance
Exclusive contracts and market dominance Giacomo Calzolari and Vincenzo Denicolò Online Appendix. Proofs for the baseline model This Section provides the proofs of Propositions and 2. Proof of Proposition.
More informationOnline Appendix for Coordination of Outsourced Operations at a Third-Party Facility Subject to Booking, Overtime, and Tardiness Costs
Submitted to Operations Research manuscript OPRE-2009-04-180 Online Appendix for Coordination of Outsourced Operations at a Third-Party Facility Subject to Booking, Overtime, and Tardiness Costs Xiaoqiang
More informationTHE HEAVY-TRAFFIC BOTTLENECK PHENOMENON IN OPEN QUEUEING NETWORKS. S. Suresh and W. Whitt AT&T Bell Laboratories Murray Hill, New Jersey 07974
THE HEAVY-TRAFFIC BOTTLENECK PHENOMENON IN OPEN QUEUEING NETWORKS by S. Suresh and W. Whitt AT&T Bell Laboratories Murray Hill, New Jersey 07974 ABSTRACT This note describes a simulation experiment involving
More informationOperations Research Letters. Joint pricing and inventory management with deterministic demand and costly price adjustment
Operations Research Letters 40 (2012) 385 389 Contents lists available at SciVerse ScienceDirect Operations Research Letters journal homepage: www.elsevier.com/locate/orl Joint pricing and inventory management
More informationLearning and Risk Aversion
Learning and Risk Aversion Carlos Oyarzun Texas A&M University Rajiv Sarin Texas A&M University January 2007 Abstract Learning describes how behavior changes in response to experience. We consider how
More information