Executive Summary : | The project proposes novel control design architectures which will enable unmanned aerial vehicles (UAVs) to perform a wide range of autonomous missions. Top-to-bottom planning hierarchy for autonomous vehicles comprises mission planning, trajectory planning, and autopilots controlling the air-frame. Conventional algorithm/control design for these planning sub-layers is decoupled and carried out independently. Typically, a mission involves various trajectory planning sub-problems with their distinct sensory and computational requirements. This presents a big bottleneck in reaching higher levels of autonomous behavior for payload limited aerial vehicles. The key idea here is to develop diverse closed-loop behavior of underlying vehicle kinematics utilizing the same sensor information and control structure. The effort essentially involves a rigorous analysis of the nonlinear vehicle kinematics, development of control architectures offering greater diversity in closed-loop response, and an extensive experimental validation of various mission scenarios. With the focus on innovative control design using the same sensor information and a wider mission paradigm, the proposal aims at breaking certain specific barriers in unifying trajectory and mission planning aspects of UAV autonomy. |