Beruflich Dokumente
Kultur Dokumente
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=informs. .
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
INFORMS is collaborating with JSTOR to digitize, preserve and extend access to Management Science.
http://www.jstor.org
An 0(T3) Algorithm for the Economic Lot-
sizing Problem with Constant Capacities
W e develop an algorithm that solves the constant capacities economic lot-sizing problem
with concave production costs and linear holding costs in o(T3) time. The algorithm is
based on the standard dynamic programming approach which requires the computation of the
minimal costs for all possible subplans of the production plan. Instead of computing these costs
in a straightforward manner, we use structural properties of optimal subplans to arrive at a
more efficient implementation. Our algorithm improves upon the o(T4) running time of an
earlier algorithm.
(EconomicLot-sizing;Complexity)
0025-1909/96/4201 /0142$01.25
Copyright ? 1996, Institute for Operations Research
142 MANAGEMENT SCIENCE/Vol. 42, No. 1, January 1996 and the Management
Sciences
VAN HOESEL AND WAGELMANS
The EconomicLot-sizingProblem
by Chen et al. (1994). We also refer to that paper for a period t - 1 without changing the set of feasible solu-
discussion of other work on NP-hard versions of the tions.
capacitated economic lot-sizing problem. (b) The unit holding costs are all equal to zero. If this
With respect to polynomially solvable special cases of is not the case, then an equivalent problem is obtained
the capacitated economic lot-sizing problem, the follow- by omitting the holding costs and redefining the pro-
ing results are known. Bitran and Yanasse showed that duction costs as pt(xt) = pt(xt) + 1T=,hix, (cf. Wagelmans
ND/Z/ND/NI and CIZICIG can be solved in O(T), et al. 1992). Note that we can achieve this only when
respectively, O(T log T) time. Chung and Lin gave an the original holding costs are linear.
O(T2) algorithm for NI/G/NI/ND and an o(T4) algo- For notational convenience, we let cf(t) denote the
rithm for C / GG / / C was presented by Florian and cost of producing at full capacity in period t, i.e., cf(t)
Klein (1971). The latter algorithm also solves the more =ft + pt(C) (t E {1, . . . , TI).
general constant capacity problem in which the cost We call production in a period t fractionalif it is be-
functions are concave instead of linear. Pochet and Wol- tween 0 and C, i.e., 0 < xt < C. Florian and Klein (1971)
sey (1993) consider the related problem in which mul- have shown that there exists an optimal schedule such
tiple batches of equal capacity are available, each re- that between any pair of fractional production periods
quiring a set-up cost. They solve this problem in o(T3) there is at least one period with zero inventory. This
time. property is often referred to as the fractionalproduction
In this paper we will show that when the production property.It also holds in case of general capacities. For
costs are concave and the holding costs are linear, it is any feasible solution, we define a subplan (t1, t2) (1 c tl
possible to solve the economic lot-sizing problem with c t2 c T) as a set of consecutive periods, starting with
constant capacities in O(T3) time. Hence, for this case t1 and ending with t2, such that at most one period has
we improve upon the Florian-Klein algorithm. fractional production and It1,- = It2= 0. (Note that our
This paper is organized as follows. In ?2 we introduce definition of subplan is more general than the usual def-
some notation. Section 3 contains a description of a inition in which inventories of intermediate periods ti,
greedy algorithm for solving a basic subproblem. In ?4 ... , t2 - 1 are required to be strictly positive.) It follows
the actual algorithm is described. Section 5 contains con- from the fractional production property that we need to
clusions and some remarks. consider only feasible solutions that can be subdivided
into subplans. This suggests an approach in which we
2. Preliminaries first determine optimal solutions for all subplans and
We will use the following notation: then choose the best combination of subplans which
T: the length of the planning horizon; constitute a complete solution. In the next section we
C: the production capacity in each period. present an algorithm for finding an optimal solution for
Furthermore, for each period t E {1, . . ., TI: a given subplan.
d,: the demand in t;
x,: the production level in t;
It: the inventory level at the end of t (Io = 0); 3. Greedy Algorithm
f,: the set-up cost in t; Consider a fixed subplan (tl, t2) for which we want to
ht: the unit holding cost in t; and find a minimum cost solution. In case dt,J2 = KC for
pt(xt): the production costs in t, a concave function some integer K, any feasible solution has only full pro-
of xt. duction-periods, namely exactly K. Hence, finding a
The cumulative demand of a set of consecutive peri- minimum cost solution for the subplan boils down to
ods Is, ..., tI (1 c s < t c T) willbedenotedbyds,t determining in which K periods full production should
=>2T=5 dT - take place. In case cumulative demand is not a multiple
Without loss of generality we may assume: of C, i.e., if dt1,J2= f + KC for some integer K and f such
(a) For each period t: dt c C. If this is not the case, that 0 < f < C, then any feasible solution has K full
we can move the excess demand in t to the preceding production periods and a fractional period in which the
Table 1 Dataof Example period T for which A(T) = k holds. The following is
obvious.
r 1 2 3 4 5 6 7
FEASIBILITY CONDITION.A production schedule is
d, 0 4 2 1 4 5 2 feasible, i.e., IT2 0 for all T E Itl, . . ., t2l, if and only if
fr 4 7 5 8 7 7 5 for every period Wk(1 c k c K), there are at least k
PI 3 1 0 1 2 1 1
production periods in Itl, .., Wkl.
cf(T) 19 12 5 13 17 12 10
This period Wkis called a choiceperiodbecause it forces
us to choose a kth full production period in the set
{1, ..., Wkl . We choose this production period as spec-
production level equals f. Suppose that we fix the frac- ified below.
tional production period, then the problem is again to GREEDYALGORITHM. Start with the production plan
determine an optimal set of full production periods. In in which only the fractional production takes place in
the remainder of this paper we will focus on the case in period t. This period is not available for full production.
which the subplan contains a fractional period, because The K full production periods are chosen as follows.
this problem is clearly at least as hard as the problem Considerthe choice periods Wk, k E 11, ..., KI, in in-
without a fractional period. creasing order. The cheapest available period T in the
We can restrict the fractional production f to periods set It,,. . ., Wkl is chosen as production period, i.e., cf(T)
t with d, 12 2 f, since fractional production in later peri- is minimal among the available periods T E tl, ... , Wkl.
ods will lead to positive ending inventory in period t2, If necessary, break ties by choosing the earliest period.
contradicting the definition of a subplan. Therefore, we EXAMPLE. We consider subplan (1, 7). The capacity
define tmaxto be the latest period t such that d, 2 2 f. C is five units. The cumulative demand is 18, and there-
Similarly, there is an earliest possible fractional period. fore K = 3 and f = 3. The other data are given in Table
If dtlt,> (t - 1)C + f, then the periods t1 through t must 1, where PTdenotes the unit production costs in period
be full capacity production periods. Therefore, we de- T, i.e., production costs are linear.
fine tmin as the first period t for which d,,c (t - 1)C Let period 4 be the fractional period, i.e., X4 = 3. Then
+ f. the calculations of the greedy algorithm are shown in
Suppose the fractional period is fixed to t e {tmin, ..., Table 2.
tmaxl and let P(t) denote the corresponding problem of The choice periods are 2, 3, and 6. In this example,
determining optimal full production periods. We intro- the full production periods coincide with the choice pe-
duce a function A(T) (T E It1, . . ., t2D)which denotes riods. However, this is not the case in general, as can be
the minimum number of full production periods in seen by swapping the cost structure of periods 1 and 2.
It1, . . . , TI in any feasible solution of P(t): This would leave the choice periods unchanged, but pe-
riod 1 would be chosen as a full production period in-
stead of period 2. Finally, we note that the total cost of
[dt1l, for T < t, this plan is 12 + 5 + (8 + 3 *1) + 12 = 40.
A(T)=
1rc
1dtl,T - ] forT ?t
Table 2 Results of GreedyAlgorithm
A solution of P(t) is feasible if and only if for any T
r 1 2 3 4 5 6 7
t1, .. ., t2l the number of full production periods in
It1,. .., T is at least A(T). The function A is integral dl,r0 4 6
and monotonically nondecreasing for the periods dl,, - f 4 8 13 15
It1, ..., t - 11 and It, ..., t2I. Moreover, A(t - 1) A(T) 0 1 2 1 2 3 3
choice n y y n n y n
c A(t) + 1. Note that A can take on the values O,... , K}.
full prod. n y y n n y n
Define for all k E {1, . . . , KI the period Wkas the earliest
The definition of choice period Wkensures that the kth When referring to the optimal solution in the remain-
full production period is chosen from a set of available der of this paper, we will mean the lexicographically
periods which is as large as possible. The greedy aspect earliest optimal solution.
of the algorithm is that among all these available peri- By solving P(t) for all t G Itmin, ... tmaxI we can
,
ods the cheapest one is chosen. Clearly, the greedy so- determine the optimal solution for subplan (tl, t2). In-
lution is feasible. Its optimality is proved next. stead of solving each of these problems separately, we
Let us first define an ordering on the feasible solutions propose an iterative algorithm in ?4. This algorithm
of P(t). Consider two feasible production plans S and computes not only the optimal solutions of the prob-
S' and the first full production period in which they lems P(t) (t E Itmin, . . . tmaxI), but also the optimal so-
,
differ. If that period is earlier in S than it is in S', then lutions of the problems defined as follows. Let t EI tmin,
solution S is called lexicographicallyearlierthan solution ... . tmaxl, then P(t)' is the problem of finding an opti-
S'. Note that the number of full production periods is mal schedule when f units become available in period t
equal to K in both solutions. completely for free, i.e., without costing any money or ca-
pacity. Clearly, a feasible solution for this problem cor-
LEMMA 1. The greedy algorithm constructs the lexico-
responds to a choice of K full production periods. The
graphicallyearliestoptimalproductionplan for P(t).
only difference with problem P(t) is that period t is now
PROOF. Let S be the lexicographically earliest opti- also available for full production (at cost cf(t)). It is eas-
mal solution. Suppose it is not equal to the solution SG ily seen that an optimal solution of P(t)' can be found
created by the greedy algorithm. by applying the greedy algorithm. Note that the choice
Let w1, . . ., WK be the choice periods for the greedy periods for P(t) and P(t)' are identical. Again, when
algorithm, and let T1, ..., TK be the respective full pro- referring to the optimal solution of P(t)', we will mean
duction periods chosen by the greedy algorithm. Let k the solution constructed by the greedy algorithm. The
be the smallest index such that Tk is not in S. There is a following properties play a key role in the algorithm.
period T' in {t1, ..., Wkl that is a production period in
LEMMA2. Lett E Itmins,. . . tmax). Theoptimalsolutions
,
S but not in SG, because otherwise S would have less
of P(t + 1)' and P(t)' differwith respectto the full produc-
than k production periods in It1, ..., Wk), violating the
tion periods in at most one period. Moreover, if there is a
feasibility condition.
difference,then the optimalsolution of P(t)' is obtainedfrom
Consider the following cases.
the optimalsolution of P(t + 1)' by movingproductionfrom
(1) If cf(T') < cf(Tk), then this contradicts the fact that
a periodin It,, . . , tI to a periodin It + 1, . , t21.
. .
the greedy algorithm chooses the cheapest available
production period for Wk. PROOF. We will prove that the solutions produced
(2) If cf(T') = Cf(Tk) and T' < Tk, then this contradicts by the greedy algorithm in both problems differ in at
the fact that the greedy algorithm chooses the earliest most one production period, as described in the lemma.
period among the cheapest available ones. The problems P(t)' and P(t + 1)' differ with respect
The feasibility condition is also satisfied by the solu- to function A only in period t:
tion created from S by replacing T' by Tk as a production
period. Therefore, we can conclude the following.
(3) If cf(T') > Cf(Tk), then the solution S can be im- A(t) =[ t] in P(t + 1)' and
proved.
(4) If cf(T') = Cf(Tk) and T' > Tk, then the solution
S is not the lexicographically earliest optimal solu- A(t) = dt1t- f] in P(t)'.
tion.
Hence, the assumption that T' * Tk always Thus, A(t) may be one unit less in P(t)' than in
leads to a contradiction. We conclude that SGis equal P(t + 1)'. This gives a possible difference in the set of
to S, the lexicographically earliest optimal solu- choice periods, which can only occur if t is a choice pe-
tion. O riod in P(t + 1)', say the kth. In that case, the kth choice
period in P(t)' may be a period u with t < u < Wk+1. equal, we have the following result, which can be
All other choice periods are identical in both problems. proved using arguments similar to those in the proof
Clearly, because the first k - 1 choice periods are of Lemma 2.
identical, the first k - 1 production periods chosen by
LEMMA3. Let t E tmax}and supposethat t is
{tmil, ...,
the greedy algorithm will be the same for both prob-
a full productionperiodin the optimalsolutionof P(t)'. Then
lems. If all choice periods are identical in both problems,
the optimalsolution of P(t) differsfrom the optimalsolution
or if the greedy algorithm chooses the same production
of P(t)' only in the fact that the full productionin t is real-
periods at t and u, then the optimal solutions do not
located.
differ. Hence, we only have to examine the case where
the choices in t and u differ, say T' is chosen at t in
problem P(t + 1)', and T" is chosen at u in prob- 4. Global Algorithm
lem P(t)'. The global algorithm for solving the lot-sizing problem
By definition, T' is the cheapest available period in consists of two phases. In the first phase we find the
It1, ..., tl, and T" is the cheapest available period in optimal solutions for the subplans. In the second phase
It1, ..., ul. Thus, T" * T' implies T" > t and cf(T") these solutions are used to determine an optimal solu-
< cf(T').
tion of the overall problem.
We show that in the remainder of the greedy algo- Phase 1: Find the minimum cost for all subplans
rithm the number of different production periods for
(tl, t2), 1 c'_ tl -'- t2 cT.
both problems remains at most one, and that the differ- Phase 2: Find, in the directed graph with vertices
ence is always as specified in the lemma. I{O,. . ., Tj and arcs (tl. - 1, t2), 1 c-- tl c-- t2 c--T, the
As argued before, the choice periods after u are equal shortest path from vertex 0 to vertex T, where the length
for both problems. Let those periods be Wk+l, . . .* WK, of arc (t, - 1, t2) is equal to the minimum cost of sub-
and consider the production period chosen at Wk+1. plan (tl, t2).
(a) Suppose that T' is the period chosen at Wk+1 in Except for vertex 0, the vertices on the shortest path
problem P(t)'. Because T' is the cheapest available pe- found in Phase 2 correspond to the last periods of the
riod up to Wk+1 in P(t)', it follows that T" is the cheapest subplans which constitute an optimal production plan.
available period up to Wk+1 in P(t + 1)'. Clearly, from Given the cost of each subplan, the second phase can be
wk+1on the partial solutions are equal again. implemented in O(T2) time, since the graph is acyclic
(b) SupposeT * T' is the period chosen at Wk+1 in and the number of arcs is O(T2). Thus, Phase 2 is not
problem P(t)'. T is the cheapest available period up to the bottleneck of the algorithm. We will therefore focus
Wk+1, and therefore T > t (since T' is the cheapest avail-
on Phase 1. By considering all possible fractional pro-
able period up to t). Moreover, in P(t + 1)' it is also the duction periods and using the greedy algorithm, a min-
cheapest available period, unless T" is cheaper. How- imum cost solution for a given subplan can be found in
ever, which of these periods is chosen does not matter. O(T2) time. Because there are O(T2) possible subplans,
In both cases the difference with respect to the partial this implies an 0(T4) algorithm for Phase 1. We will
solution of P(t)' remains one period, either T or T", and give improvements that lead to an o(T3) implementa-
both are later than t. tion.
If (a) occurs, then it follows immediately that the full
production periods of the optimal solutions of P(t)' and 4.1. Iterative Algorithm for Phase 1
P(t + 1)' are equal. If (b) occurs, the above argument We show that the minimum cost of each subplan (tl, t2)
can be repeated for the later choice periods Wk+2, (1 c tl c t2 c T) can be calculated in O(T) amortized
WK, and the lemma is proved. LII time. The algorithm consists of the following steps for
If t is not chosen as a full production period in the each subplan.
optimal solution of P(t)', then it is clearly optimal Let tmin, tmax and the optimization problems P(t)' and
to take the same full production periods as in the P(t) (t E {tmin, ... tmax)) be as defined in the previous
solution of P(t). In case the optimal solutions are not section.
T 1 2 3 4 5 6 7 T 1 2 3 4 5 6 7
Cf(T) 19 12 5 13 17 12 10 Cf(T) 19 12 5 13 17 12 10
prod. n y y n y n, n prod. n y y n n y n
IT 0 1 4 3 7 2 0 IT 0 1 4 3 2 2 0
6, 1 1 1 4 4 6 6, 1 1 1 4 4
7YT 5
6T 1 1 1 4 4
timal solution of P(t)', then the optimal solution of P(t)
similar subproblems have to be solved. The algorithm Chung, C.-S. and C.-H. M. Lin, "An O(T2) Algorithm for the NI/CG
NI/ND Capacitated Lot Size Problem," Management Sci., 34
exploits the fact that the optimal solutions to these prob-
(1988), 420-426.
lems are partially equal. The only possible way in which Federgruen, A. and M. Tzur, "A Simple Forward Algorithm to Solve
a further improvement could be achieved seems to be General Dynamic Lot Sizing Models with n Periods in 0(n log n)
the exploitation of relations concerning the positioning or 0(n) Time," ManagementSci., 37 (1991), 909-925.
of optimal fractional periods in closely related subplans. Florian, M. and M. Klein, "Deterministic Production Planning with
Until now, we have not been able to identify such re- Concave Costs and Capacity Constraints," ManagementSci., 18
(1971), 12-20.
lations.'
, J. K. Lenstra, and A. H. G. Rinnooy Kan, "Deterministic Produc-
tion Planning: Algorithms and Complexity," ManagementSci., 26
1 The authors would like to thank two anonymous referees and an (1980), 669-679.
associate editor of ManagementSciencefor their comments on an earlier Van Hoesel, S., A. Wagelmans, and B. Moerman, "Using Geometric
draft of this paper. Techniques to Improve Dynamic Programming Algorithms for
the Economic Lot-Sizing Problem and Extensions," EuropeanJ.
References Oper. Res., 75 (1994), 312-331.
Aggarwal, A. and J. K. Park, "Improved Algorithms for Economic Lot- Pochet, Y. and L. A. Wolsey, "Lot-Sizing with Constant Batches: For-
Size Problems," Oper. Res., 41, 3 (1993), 549-571. mulation and Valid Inequalities," Mathematicsof Oper. Res., 18
Bitran, G. B. and H. H. Yanasse, "Computational Complexity of the (1993), 767-785.
Capacitated Lot Size Problem," ManagementSci., 28, 10 (1982), Veinott, A. F., Jr., Unpublished class notes, Program in Operations
1174-1185. Research, Stanford University, Stanford, California, 1963.
Chen, H.-D., D. W. Hearn, and C.-Y. Lee, "A New Dynamic Program- Wagelmans, A., S. Van Hoesel, and A. Kolen, "Economic Lot-Sizing:
ming Algorithm for the Single Item Capacitated Dynamic Lot Size An 0(n log n) Algorithm that Runs in Linear Time in the Wagner-
Model," J. of GlobalOptimization,4 (1994), 285-300. Whitin Case, Oper. Res., 40 (1992), S145-156.
Acceptedby Luk Van Wassenhove;receivedJuly 1993. This paperhas beenwith the authors5 monthsfor 1 revision.