Calculating flight time for unmanned aerial vehicles in the presence of obstacles and the incorporation of flight dynamics
Myers, David J.
MetadataShow full item record
With the continuously growing use of unmanned aerial vehicles in many domains, efficient routing is very important. The need for accurate flight time data is evident in these algorithms, and current limiting assumptions used for the calculation of flight times can be relaxed or eliminated in order to improve the flight time data. We use a network generation procedure to allow obstacles to be placed in the operational field and to include flight dynamics in the flight time calculations. The network is composed of bases, tasks, and obstacle defining vertices. Pseudonodes and associated edges are added into the network in order to include flight dynamics. A selective Dijkstra's shortest path method was designed and implemented in order to ensure that the shortest path between a vertex pair of interest includes flight dynamics. A vertex pair of interest is any pair of task/base, base/base, or task/task vertices in the operational field. A large operational field containing thirty tasks and three obstacles would be considered a complex mission plan. Our network generation procedure along with shortest path calculations for all 992 vertex pairs of interest solves in approximately one second. This procedure allows for the fast computation needed for a dynamic domain such as the use of mission planning algorithms and dynamic reassignment algorithms for military use of unmanned aerial vehicles.