A hierarchical/intelligent control architecture for unmanned aerial vehicles

In this paper, a hierarchical/intelligent control architecture for an unmanned aerial vehicle (UAV) is proposed. The architecture consists of three levels: the highest level is occupied by mission planning routines. At this level, information about the way points the vehicle must follow is available and logic-based routines decide upon mission tasks while maintaining physical constraints and generate the task queue. The mid-level controller coordinates the task execution while a trajectory planning component receives the task information from the high-level module and provides set points for low-level stabilizing controllers whose function is to maintain the vehicle in a stable state and to follow accurately the commanded trajectory. An adaptive mode transitioning control algorithm resides also at the lowest level of the hierarchy consisting of two components: a mode transitioning controller and the accompanying adaptation mechanism. The adaptation routine may be turned on only when needed. The transitioning algorithm operates in real-time while adapting on-line to disturbances and other external inputs. This intelligent/hierarchical architecture is being implemented using a novel software infrastructure called Open Control Platform, which facilitates interoperability, plug-and-play and other functionalities. Simulation results illustrate the robustness and effectiveness of the proposed scheme. An actual flight demonstration is planned for the near future as part of a DARPA sponsored research program.