Dynamic grouping of cooperating vehicles using a receding horizon controller for ground target search and track missions
Cameron K. Peterson · 2017 IEEE Conference on Control Technology and Applications (CCTA) · 2017
Teams of unmanned vehicles are capable of accomplishing a wide variety of mission objectives, such as searching for and tracking targets. In this paper, a receding horizon control is utilized with information based reward measures to accomplish these two competing mission objectives. This approach for cooperatively searching and tracking has proven to be effective in past work. However, it is not generally scalable for large numbers of vehicles due to the computational expense required when generating joint path decisions. This paper proposes a method to dynamically group vehicles with neighbors that have intersecting decision spaces, thus reducing computational cost while still maintaining reasonable performance. Each vehicle also decides its ideal event horizon based upon inferred knowledge of the operational environment, further reducing cost.