 Original Paper
 Open Access
A time expanded version of the (Hitchcock) transportation problem
 Raffaele Mosca^{1}Email author
https://doi.org/10.1007/s1254401100563
© The Author(s) 2011
 Received: 27 January 2011
 Accepted: 7 September 2011
 Published: 1 October 2011
Abstract
This paper considers a time expanded version of the classical (Hitchcock) transportation problem. Certain quantities of goods are produced in the production facilities in each period of a planning horizon and also the demand values of the outlets are different for each time period. Moreover, the transportation costs from i to j also vary over time. The different planning periods are connected by the fact that goods may be stored at a facility or at an outlet to exploit cheaper transportation costs as long as the demand is met in every time period. We describe the problem by two models, showing that it can be solved in polynomial time by standard software packages. Then we describe some generalizations of the problem by adapting the two models, pointing out that each model seems to provide different benefits.
Keywords
 Transportation problems
 Mincost flow problem
 Hitchcock problem
1 Introduction
For notation and standard terminology, let us refer to [10] (see also [6, 9]).
The (Hitchcock) transportation problem is a well known problem in operations research [6, 9, 10] and can be solved in polynomial time by fast algorithms (see e.g. [12]). By its applicative nature, several versions of this problem have been considered, such as dynamic, timedependent versions (see e.g. [5]).
In this paper we consider a time expanded version, which is defined below as problem P. Problem P was studied in [2] (see also [3, 4]) by referring to a method for solving integer programming which resorts to the calculation of suitable Gröbner bases [7]: in particular, the authors call it 3dimensional transportation problem (referring to [11], while a similar name was used in [1] for a different problem). The motivation of this paper is to try to study problem P by an operations research approach. The closest reference in this sense seems to be [8], where the authors study the existence of feasible solutions in a multiperiod allocation problem of substitutable resources: in particular, at a certain point of the paper the authors call it a multiperiod transportation problem a problem which however seems to be quite distant from problem P.
Let P be the following problem: Given r production facilities F_{1}, ..., F_{ r }, let A_{ ik } (i = 1, ..., r, and k = 1, ..., t) denote the number of units of an indivisible good produced by F_{ i } during the kth period of a planning horizon of t periods. Assume that there are s outlets O_{1}, ..., O_{ s } each one demanding a certain number of units per period, say B_{ jk } (j = 1, ..., s and k : = 1, ..., t). Let c_{ ijk } ≥ 0 be the cost associated with transporting one unit from F_{ i } to O_{ j } during the kth period, let h_{ ik } ≥ 0 (let h′_{ jk } ≥ 0) be the cost associated with storing one unit in F_{ i } (in O_{ j }) at the end of the kth period. Then one wishes to minimize the total cost of transportation and of storage along the planning horizon of t periods.

x_{ ijk } the amount of good which is sent from F_{ i } to O_{ j } in the kth period;

z_{ ik } the amount of good which is stored in F_{ i } at the end of the kth period;

z′_{ jk } the amount of good which is stored in O_{ j } at the end of the kth period.
In this paper we propose two models for P: the first (Section 2) is a natural mincost flow model, while the second (Section 3) is a Hitchcock model obtained by exploiting some peculiarities of the problem. Then we describe some possible generalizations of P by adapting the two proposed models (Section 4), pointing out that each model seems to provide different benefits.
2 A mincost flow model for P
 (i)
0 ≤ f(u, v) ≤ w(u, v) for all (u, v) ∈ E
 (ii)
∑ _{(u,v) ∈ E}f(u, v) = ∑ _{(v,u) ∈ E}f(v, u) for all v ∈ V ∖ {σ, τ}.
Given a network N and a flowvalue v_{ f }, the mincost flow problem with respect to pair (N, v_{ f }) is that of computing in N a flow f of value v_{ f } that has minimum cost.

construct t disjoint digraphs G_{1}, ..., G_{ t } such that:whereComment: vertex a_{ ik } represents F_{ i } at period k, for i = 1, ..., r, k = 1, ..., t; vertex b_{ ik } represents O_{ j } at the period k, for j = 1, ..., s, k = 1, ..., t; edge (a_{ ik }, b_{ jk }) represents the possibility of transporting good from F_{ i } to O_{ j } at period k, for i = 1, ..., r, j = 1, ..., s, k = 1, ..., t.

add vertex σ and edges {(σ, a_{ ik }) : i = 1, ..., r, k = 1, ..., t};

add vertex τ and edges {(b_{ jk }, τ) : j = 1, ..., s, k = 1, ..., t};

add edges {(a_{ ik }, a_{ik + 1}) : i = 1, ..., r, k = 1, ..., t − 1};

add edges {(b_{ jk }, b_{jk + 1}) : j = 1, ..., s, k = 1, ..., t − 1};

the elements of V and E are defined as above;

x_{ ijk } = y(a_{ ik }, b_{ jk }) for i = 1, ..., r, j = 1, ..., s, k = 1, ..., t;

z_{ ik } = y(a_{ ik }, a_{ik + 1}) for i = 1, ..., r, k = 1, ..., t;

z′_{ jk } = y(b_{ jk }, b_{jk + 1}) for j = 1, ..., s, k = 1, ..., t.
One can formalize what above by the following theorem.
Theorem 1
An instance of P can be solved as a mincost flow problem.
As network N can be efficiently constructed, one obtains the following corollary.
Corollary 1
Problem P can be solved in polynomial time.
In particular the problem admits an optimal integer solution if all the values A_{ ik }, B_{ jk } are integer.
3 A Hitchcock model for P
Let us consider problem P as problem P_{ f } defined in Section 2, according to the equalities linking the respective solutions. In [10] a standard transformation from an instance of the mincost flow problem to an instance of the Hitchcock problem is shown. However, in the sequel let us try to model P as a Hitchcock problem just by exploiting some peculiarities of the problem (i.e., by compacting some aspects of P_{ f }) and not by using the standard transformation of [10].
Assume that i′ > j′. Then one can define c_{ii′jj′} = ∞, so to ensure q_{ii′jj′} = 0 for i′ > j′.
Assume that i′ ≤ j′.
Let us observe that for any solution of P, the amount q_{ii′jj′} will arrive from F_{ i } to O_{ j } through a path from a_{ii′} to b_{jj′} in the network N of Section 2: in particular, all the edgecapacities along each of such paths are ∞.
It follows that for any optimal solution of P, the amount q_{ii′jj′} will arrive from F_{ i } to O_{ j } through a least cost path (or in general through least cost paths, if more than one) from a_{ii′} to b_{jj′} in the network N of Section 2. Then the cost c_{ii′jj′} may be defined as the cost of a least cost path from a_{ii′} to b_{jj′}: by the structure of N such paths are exactly j′ − i′ + 1 (i.e., each such path is formed by a possible storage multi period in F_{ i }, by a transfer period from F_{ i } to O_{ j }, and by a possible storage multi period in O_{ j }: the transfer period is a period k with i′ ≤ k ≤ j′).

c_{ii′jj′} = ∞, for i′ > j′;

c_{ii′jj′} = min_{i′ ≤ k ≤ j′} {( ∑ _{u = i′,...,k − 1}h_{ iu }) + c_{ ijk } + ( ∑ _{u = k,...,j′}h_{ ju })}, for i′ ≤ j′.
Let us show that given an optimal solution [x_{ ijk }, z_{ ik }, z′_{ jk }] of P one can derive a feasible solution [q_{ii′jj′}] of H, and viceversa.
From P to H Let [x_{ ijk }, z_{ ik }, z′_{ jk }] be an optimal solution of P. Then a solution [q_{ii′jj′}] of H with the same cost can be obtained as follows. Let us observe that: (i) the costs of storage in facilities (in outlet) do not depend on the destination (on the origin) of the good. By (i) and by the definition of costs c_{ii′jj′}, one may proceed as follows.
The first step may be to obtain the values of q_{ ikjk }, for k = 1, ..., t: to this end, for k = 1, ..., t, it is enough to chose the values of q_{ ikjk }, in order to maximize ∑ _{i = 1,...,r} ∑ _{j = 1,...,s}q_{ ikjk }, subject to q_{ ikjk } ≤ x_{ ijk } and to ∑ _{j = 1,...,s}q_{ ikjk } ≤ A_{ ik } for i = 1, ..., r (that is to charge the values of x_{ ijk } to those of q_{ ikjk } as much as possible). That can be easily done by a greedy technique.

w((σ, a_{ ik })) = A_{ ik } − ∑ _{j = 1,...,s}q_{ ikjk }, for i = 1, ..., r, k = 1, ..., t;

w((b_{ jk }, τ)) = B_{ jk } − ∑ _{i = 1,...,r}q_{ ikjk }, for j = 1, ..., s, k = 1, ..., t;

w((a_{ ik },a_{ik + 1})) = z_{ ik }, for i = 1, ..., r, k = 1, ..., t;

w((b_{ ik }, b_{ik + 1})) = z′_{ jk }, for j = 1, ..., s, k = 1, ..., t;

w((a_{ ik }, b_{ ik })) = x_{ ijk } − q_{ ikjk }, for i = 1, ..., r, j = 1, ..., s, k = 1, ..., t.
From H to P Let [q_{ii′jj′}] be a solution of H. Then a solution [x_{ ijk }, z_{ ik }, z′_{ jk }] of P with the same cost can be obtained as follows. By the definition of costs c_{ii′jj′}, every q_{ii′jj′} defines a least cost path P_{ii′jj′} from a_{ii′} to b_{jj′} in the network N of Section 2 (i.e., corresponding to the costs which minimize the expression which defines c_{ii′jj′}). Then q_{ii′jj′} can be charged (as a flow line) onto the edges of P_{ii′jj′}, so to compose the values of a solution [x_{ ijk }, z_{ ik }, z′_{ jk }] of P according to the equalities of Section 2.
One can formalize what above by the following theorem.
Theorem 2
An instance of P can be solved (directly) as a Hitchcock problem.
Then a corollary and a comment similar to those at the end of Section 2 hold.
4 Some generalizations of P
In this section let us introduce three possible generalizations of problem P, pointing out the advantages and the disadvantage of the two proposed models.
Let P_{1} be the following problem. In the context of problem P, assume that every facility F_{ i } for i = 1, ..., r (that every outlet O_{ j } for j = 1, ..., s) can store at most d_{ i } (at most d′_{ j }) units of good at the end of each period k for k = 1, ..., t. The objective remains that of the problem P.

w((a_{ ik }, a_{ik + 1})) = d_{ i }, for i = 1, ..., r, k = 1, ..., t;

w((b_{ jk }, b_{jk + 1})) = d′_{ j }, for j = 1, ..., s, k = 1, ..., t.
On the other hand, it seems to be not immediate to model P_{1} by the Hitchcock model of Section 3.
Let P_{2} be the following problem. In the context of problem P, assume that the good produced in a facility has to be sold in a outlet before a certain number of periods (e.g., before corruption of the good), say L periods. The objective remains that of the problem P.

, for k = 1, ..., t, where k^{*} = min{k + L, s}.

c_{ii′jj′} = ∞ , for j′ − i′ > L,
Then P_{2} can be solved as a Hitchcock problem, i.e., in polynomial time.
Let P_{3} be the following problem: In the context of problem P, let c_{ ik } (i = 1, ..., r, k = 1, ..., t) be the cost of production for a unit of good in F_{ i } in the kth period, and let p_{ jk } (j = 1, ..., s, k = 1, ..., t) be the profit for selling a unit of good in O_{ j } in the kth period: in particular, A_{ ik } can be view as the maximum amount which can be produced by F_{ i } in the kth period. Then one wishes to minimize the total cost of transportation, of storage, and of production, less the total profit by selling, along the horizon of t periods, i.e., one wishes to maximize the total profit by selling, less the total cost of transportation, of storage, and of production, along the horizon of t periods.
A possible scenario for problem P_{3} may be that of a tobacco manufacture company: in fact the selling price of tobacco varies over countries and its fluctuation (over countries) is planned in advance.

c((σ, a_{ ik })) = c_{ ik }, for i = 1, ..., r, k = 1, ..., t;

c((σ, b_{ jk })) = − p_{ jk }, for j = 1, ..., s, k = 1, ..., t.
Then P_{3} can be modeled by the Hitchcock model of Section 3 as follows.

i′ ≤ j′ (it is possible);

c_{ii′} +min_{i′ ≤ k ≤ j′} {( ∑ _{u = i′,...,k − 1}h_{ iu }) + c_{ ijk } + ( ∑ _{u = k,...,j′}h_{ ju })} − p_{jj′} < 0 (it is convenient).
For any green variable q_{ii′jj′} let us write c_{ii′} + min_{i′ ≤ k ≤ j′} {( ∑ _{u = i′,...,k − 1}h_{ iu }) + c_{ ijk } + ( ∑ _{u = k,...,j′}h_{ ju })} − p_{jj′}, and let K be a scalar with .

, if q_{ii′jj′} is a green variable;

, otherwise.

, if q_{ii′jj′} is a green variable;

q_{ii′jj′} = 0, otherwise.
Then summarizing it seems that: the first model is more powerful than the second model to describe generalizations of P (e.g., P_{1}), though it may lead to linear programming which can not be solved as a classical mincost flow problem (e.g., P_{2} and P_{3}); while the second model may be more useful than the first model to show that certain generalizations of P (e.g., P_{2} and P_{3}) can be solved as a classical Hitchcock problem.
5 Conclusions
The (Hitchcock) transportation problem is a well known problem in operations research [6, 9, 10] and can be solved in polynomial time by fast algorithms (see e.g. [12]). By its applicative nature, several versions of this problem have been considered, such as dynamic, timedependent versions (see e.g. [5]).
In particular the following time dependent version was studied in [2] (see also [3, 4]) by referring to a method for solving integer programming which resorts to the calculation of suitable Gröbner bases [7]:
Given r production facilities F_{1}, ..., F_{ r }, let A_{ ik } (i = 1, ..., r, and k = 1, ..., t) denote the number of units of an indivisible good produced by F_{ i } during the kth period of a planning horizon of t periods. Assume that there are s outlets O_{1}, ..., O_{ s } each one demanding a certain number of units per period, say B_{ jk } (j = 1, ..., s and k : = 1, ..., t). Let c_{ ijk } ≥ 0 be the cost associated with transporting one unit from F_{ i } to O_{ j } during the kth period, let h_{ ik } ≥ 0 (let h′_{ jk } ≥ 0) be the cost associated with storing one unit in F_{ i } (in O_{ j }) at the end of the kth period. Then one wishes to minimize the total cost of transportation and of storage along the planning horizon of t periods.
In this paper we tried to study the above problem by an operations research approach. We described the problem by two models, showing that it can be solved in polynomial time by standard software packages: the first is a natural mincost flow model, while the second is a Hitchcock model obtained by exploiting some peculiarities of the problem. Then we described some possible generalizations of the problem by adapting the two proposed models, pointing out that each model seems to provide different benefits.
Declarations
Acknowledgements
I would like to thank Prof. Giandomenico Boffi and Prof. Fabio Rossi for having kindly explained their work [2] in 2003. Then would like to thank the referees for their comments. Finally would like humbly to dedicate this paper to (the memory of) my grandfather Raffaele Mosca.
Authors’ Affiliations
References
 Bein W, Brucker P, Park JK, Pathak PK (1995) A Monge property for the ddimensional transportation problem. Discrete Appl Math 58:97–109MATHMathSciNetView ArticleGoogle Scholar
 Boffi G, Rossi F (2000) Gröbner bases related to 3dimensional transportation problems. In: Quaderni matematici dell’ Universitá di Trieste, vol 482. DSMTrieste. http://www.dmi.units.it/~rossif/
 Boffi G, Rossi F (2001) Lexicographic Gröbner bases of 3dimensional transportation problems. In: Symbolic computation: solving equations in algebra, geometry, and engineering (South Hadley, MA, 2000). Contemporary Mathematics, vol 286. Amer. Math. Soc., Providence, pp 145–168View ArticleGoogle Scholar
 Boffi G, Rossi F (2006) Lexicographic Gröbner bases for transportation problems of format r ×3 ×3. J Symb Comput 41:336–356MATHMathSciNetView ArticleGoogle Scholar
 Bookbinder JH, Sethi SP (1980) The dynamic transportation problem: a survey. Nav Res Logist Q 27:65–87MATHMathSciNetView ArticleGoogle Scholar
 Cook WJ, Cunningham WH, Pulleyblank WR, Schrijver A (1998) Combinatorial optimization. Wiley, New YorkMATHGoogle Scholar
 Hosten S, Thomas R (1998) Gröbner bases and integer programming. In: Bichberger, Winkler (eds) Gröbner bases and applications, vol 251. Cambridge University Press, Cambridge, pp 144–158View ArticleGoogle Scholar
 Klein RS, Luss H, Rothblum UG (1995) Multiperiod allocation of substitutable resourses. Eur J Oper Res 85:488–503MATHView ArticleGoogle Scholar
 Lawler E (2001) Combinatorial optimization: networks and matroids. Dover, New YorkGoogle Scholar
 Papadimitriou CH, Steiglitz K (1998) Combinatorial optimization: algorithms and complexity. Dover, New YorkMATHGoogle Scholar
 Sturmfels B (1995) Gröbner bases and convex polytopes. AMS, ProvidenceView ArticleGoogle Scholar
 Tokuyama T, Nakano J (1995) Efficient algorithms for the Hitchcock transportation problem. SIAM J Comput 24:563–578MATHMathSciNetView ArticleGoogle Scholar
Copyright
This article is published under license to BioMed Central Ltd. Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution and reproduction in any medium, provided the original author(s) and source are credited.