Ramified optimal transportation with payoff on the boundary

Abstract

This paper studies a variant of ramified/branched optimal transportation problems. Given the distributions of production capacities and market sizes, a firm looks for an allocation of productions over factories, a distribution of sales across markets, and a transport path that delivers the product to maximize its profit. Mathematically, given any two measures $\mu$ and $\nu$ on $X$, and a payoff function $h$, the planner wants to minimize $\mathbf{M}_{\alpha }(T)-\int_{X}hd(\partial T)$ among all transport paths $T$ from $\tilde{\mu}$ to $\tilde{\nu}$ with $\tilde{\mu}\leq \mu $ and $\tilde{\nu}\leq \nu $, where $\mathbf{M}_{\alpha }$ is the standard cost functional used in ramified transportation. After proving the existence result, we provide a characterization of the boundary measures of the optimal solution. They turn out to be the original measures restricted on some Borel subsets up to a Delta mass on each connected component. Our analysis further finds that as the boundary payoff increases, the corresponding solution of the current problem converges to an optimal transport path, which is the solution of the standard ramified transportation.