Akademik Veri Yönetim Sistemi
2009 IEEE Aerospace Conference, Montana, Amerika Birleşik Devletleri, 7 - 14 Mart 2009, ss.2971-2973
This paper outlines an agent based approach for UAV (Unmanned Air Vehicles) mission route planning problem. In this context, mission route planning can be defined as finding the "best" set of waypoints for the UAV that will enhance its probability of success in its mission. Mission route planning may be carried out either in premission or in mission time. It may be done either by an onboard control system or by a ground mission planning system. Regardless of these considerations, a problem solving agent for UAV mission route planning problem is proposed in this research. Mission Route Planning Agent in UAV mission route planning context needs to perceive the elements in its environment and comprehend the meaning of this situation to compute a route. In other words, it will use target, threat, terrain and air space restrictions to compute the "best" route for that UAV mission. In a dynamic re-planning environment, it will compute its new route again and again each time the situation has changed. To compute the "best" route, an A* based graph search algorithm is used. For a node, while real cost is computed using fuel and threat costs with determined weightings, heuristic cost is selected as the geometric distance. Network is constructed in the search time. Consecutive nodes are determined by spanning range and azimuth with determined increments. The values of weightings for fuel and threat costs and the increments used to construct the network have a great effect on the resulting route and the computation time. While common practice is to use default or user defined values for those, this paper presents a situation aware UAV Mission Route Planning Agent that uses the tactical situation to determine these values.(12)