Autonomous Navigation and Nonlinear Control for Quadrotors in a Structured Environment

This paper deals with the autonomous flight problem for a quadrotor. The quadrotor must reach its mission waypoints to complete a package delivery task autonomously while avoiding obstructing obstacles. A new linear matrix inequality-based dynamic gain adaptive robust control approach is proposed for the attitude and position controller of the quadrotor. The transient and steady-state tracking performances are guaranteed under bounded uncertain parameters of the quadrotor and external disturbances. To realize a higher level of automation for the quadrotor, an autonomous navigation strategy that consists of a trajectory generation approach and a sampling-based multicriteria local waypoint selection algorithm is presented. The trajectory generation approach is used to realize waypoint-based navigation with the proposed controller for the quadrotor. The sampling-based multicriteria local waypoint selection approach is proposed based on a deterministic sampling process and solved by a multicriteria optimization problem to balance path safety and speed in a structured environment. This autonomous navigation strategy enables the avoidance of path/trajectory planning, which could be time consuming in a cluttered environment. The simulation results demonstrate that the proposed controller and autonomous navigation strategy are useful for a variety of quadrotor tasks, including precise trajectory tracking and autonomous navigation in an unknown obstacle-laden environment.