Monte Carlo Tree Search for Multi-function Radar Task Scheduling
Résumé
Multi-function radars require efficient resource management strategies to fulfill their missions. In particular, task scheduling is crucial to mitigate difficult situations such as when all tasks cannot be accomplished. However, current approaches may prove insufficient in the face of emerging threats. In this article, we present a new formulation for the scheduling problem. Our model allows building schedules in a flexible way, which facilitates the discovery of high-value solutions using heuristics or tree search. We show that our algorithms provide noticeable performance improvement over similar methods proposed previously.
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|