Occupancy Based Map Searching Using Heterogeneous Teams of Autonomous Vehicles
Christopher W. Lum, Rolf T. Rysdyk, Anawat Pongpunwattana · AIAA Guidance, Navigation, and Control Conference and Exhibit · 2006
In typical search missions, the environment and the targets are not stationary and previous observations become less reliable as time progresses. In addition, the search is often initiated with only a rough idea of target location. In this work we consider a strategy for searching using a team of heterogeneous autonomous vehicles. The team members maintain a world model which includes the estimate of possible target states. The issue of compelling agents to converge on targets and to search unexplored regions is formulated as a model predictive control problem. The world model is propagated in time and strategic decisions are made autonomously based on its prediction. Agents formulate control decisions by optimizing an objective function which allows for control and timing constraints. Individual agents in the team are coupled to one another through the centralized occupancy based map. This coupling, in combination with the outlined search strategy, leads to an efficient, autonomous, and cooperative search. Nomenclature B Set of z values defining spatial domain of occupancy based map B ̃ Set of z values defining center of occupancy based map cells