Abstract | ||
---|---|---|
Patrolling is the most commonly requested type of unmanned aerial vehicle (UAV) mission; thus we use it as a prototypical example of autonomous mission planning. In this paper, we design an autonomous mission planner for UAV patrolling missions in unpredictably dynamic environments (UDEs) using recomposable restricted finite state machines (ReRFSM). UDEs admit online changes to the number of waypoints, their priorities, and the availability of paths without any prediction. ReRFSM can model UDEs and restricted inputs. Dynamic programming is used to generate policies and limited breadth-first search (LBFS) is used to generate solutions with a small computation time that plan single UAV tours. Finally, we introduce patrolling policies for multiple UAVs by composing multiple mission RFSMs and modifying the inputs of the ReRFSMs. We demonstrate the case of two vehicles as a proof of concept. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1142/S2301385017500108 | UNMANNED SYSTEMS |
Keywords | DocType | Volume |
UAV,patrolling,dynamic environments | Journal | 5 |
Issue | ISSN | Citations |
4 | 2301-3850 | 0 |
PageRank | References | Authors |
0.34 | 0 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jinwoo Seok | 1 | 0 | 0.34 |
Mariam Faied | 2 | 14 | 4.67 |
Anouck Girard | 3 | 0 | 0.34 |