Lagrangian, Eulerian and Kantorovich formulations of multi-agent optimal control problems: equivalence and gamma-convergence (Q2122150)

This paper is devoted to the study of multi-agent deterministic optimal control problems from different points of view or with different formulations. The authors call the three considered formulations Lagrangian, Eulerian and Kantorovich. As the authors say, the Lagrangian and Eulerian terminologies come from fluid-dynamics and they have been recently adopted in the theory of optimal transport, from which they also took inspiration for the Kantorovich formulation. In general terms, the Lagrangian approach consists in labeling each particle and following the corresponding trajectory. The Eulerian description, on the other hand, aims at measuring the velocity of particles flowing at a point at a fixed time. The third formulation turns out to be fundamental in connecting the Lagrangian and Eulerian points of view and it is based on the representation of solutions of the continuity equation provided by the superposition principle. For the Lagrangian formulation, the authors present also a relaxed version. They make a comparison among the four formulation showing equivalence. In the last part of this long paper, the authors model the evolution of a finite numbers of particles and study the asymptotic behaviour, when the number of particles goes to infinity, with respect to \(\Gamma\)-convergence of the Lagrangian and Eulerian cost functionals.