A hybrid control framework for an optimal visiting problem

TL;DR

This paper introduces a hybrid control framework for the optimal visiting problem, combining continuous trajectories with discrete target states, and reformulates it as a collection of optimal stopping problems to ensure well-posedness.

Contribution

It presents a novel hybrid control approach that integrates discrete target states into trajectory optimization, addressing issues where dynamic programming principles do not hold.

Findings

Well-posedness of the Hamilton-Jacobi formulation established

Reformulation as optimal stopping problems demonstrated

Hybrid approach effectively models target visiting trajectories

Abstract

The optimal visiting problem is the optimization of a trajectory that has to touch or pass as close as possible to a collection of target points. The problem does not verify the dynamic programming principle, and it needs a specific formulation to keep track of the visited target points. In this paper, we introduce a hybrid approach by adding a discontinuous part of the trajectory switching between a group of discrete states related to the targets. Then, we show the well-posedness of the related Hamilton-Jacobi problem, by reformulating the optimal visiting as a collection of time-dependent optimal stopping problems.