# Weber problem

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

A problem formulated in [a7] as a model for the optimum location of a facility in the plane intended to serve several users; for example, a central source of electric power. One is seeking the minimum of a function where the are positive scalars, the are given vectors in , is in and is the Euclidean norm of . The case when all and had been considered by P. Fermat in 1629, by E. Torricelli in 1644 and by J. Steiner in 1837. (For the early history of the problem, see [a4].)

The function is convex (cf. Convex function (of a real variable)) and one shows that, with some exceptions, it has a unique minimizer. These assertions remain valid when is allowed to be an arbitrary norm and is replaced by .

For applications, of which there are many (see [a1]), one seeks good computational methods for finding a minimizer of , either with the Euclidean norm or with other norms. For the Euclidean case, E. Weiszfeld [a8] provided a much used method; see [a6] for a discussion of this and other cases. If is the -norm, explicit solution is possible (see [a2], Chap. 4).

Minimizing the function is a problem in optimization and it is natural to seek a dual problem (cf. Duality in extremal problems and convex analysis): to maximize a function such that . A dual was found for special cases by H.W. Kuhn and others [a4]. A major result in this direction was provided by C. Witzgall in [a9], who provided a dual for a more general minimum problem, in which the function to be minimized has the form where is now allowed to be an asymmetric norm in (that is, is required to be valid only for non-negative ) and . Witzgall's result and others are subsumed under a duality theorem of W. Kaplan and W.H. Yang [a3]. In this theorem, the function has the form in , where is a norm, allowed to be asymmetric, in , is a constant -matrix, in and in are constant vectors. The dual function is the linear function in , subject to the constraints and , where and are dual norms in : . It is assumed that the equation has a solution with . In [a3] it is shown how, when the norms are differentiable (except at the origin), a minimizer of can be obtained from or determines a maximizer of . It is also shown that the theorem provides a dual for the multi-facility location problem, for which the function to be minimized is the sum of the weighted distances from new facilities to given facilities as well as the weighted distances between the new facilities; the function is a convex function of , where the th new facility is placed at .

The Weber problem has been generalized in many ways to fit the great variety of problems arising in the location of facilities. See [a1] for an overview.

How to Cite This Entry:
Weber problem. Encyclopedia of Mathematics. URL: http://encyclopediaofmath.org/index.php?title=Weber_problem&oldid=16979
This article was adapted from an original article by Wilfred Kaplan (originator), which appeared in Encyclopedia of Mathematics - ISBN 1402006098. See original article