Partial Monge–Kantorovich problem: variational formulation and numerical approximation
John W. Barrett
Imperial College London, United KingdomLeonid Prigozhin
Ben Gurion University of the Negev, Beer-Sheba, Israel
Abstract
We consider the Monge–Kantorovich problem with transportation cost equal to distance and a relaxed mass balance condition: instead of optimally transporting one given distribution of mass onto another with the same total mass, only a given amount of mass, m, has to be optimally transported. In this partial problem the given distributions are allowed to have different total masses and m should not exceed the least of them. We derive and analyze a variational formulation of the arising free boundary problem in optimal transportation. Furthermore, we introduce and analyse the finite element approximation of this formulation using the lowest order Raviart–Thomas element. Finally, we present some numerical experiments where both approximations to the optimal transportation domains and the optimal transport between them are computed.
Cite this article
John W. Barrett, Leonid Prigozhin, Partial Monge–Kantorovich problem: variational formulation and numerical approximation. Interfaces Free Bound. 11 (2009), no. 2, pp. 201–238
DOI 10.4171/IFB/209