Abstract:
In the included work the Unmanned Aerial Vehicle (UAV) mission is represented by
energy graphs motivated by the analysis in [1]. The problem of the shortest path
routing is revisited when a dynamically changing environment is considered. It
is assumed that information about the map is received while on flight due to
events. In addition, UAVs are required, while on mission, to "scout" areas of
interest which involves extracting as much intelligence as possible and
traversing it in the most safe flyable means. Hence, the UAV should be capable
of integrating knowledge from a variety of sources and re-plan its mission
accordingly in order to fulfil objectives. Motivated by the previous, depending
on the decision making process, the notion of a "temporary" optimum path can be
of physical and functional sense. The problem is modeled as a multistage
decision making process, where each stage is triggered by an event and is
characterized by a current starting point, an area for reconnaissance purposes
and a final destination. Hence, given the current availability between paths,
the objective is to devise a policy that leads from an origin or current known
location to a destination node while traversing the unknown region of interest
with the minimal energy demand.
