2016 · 96 citations · 25 references
EngineeringMechanical EngineeringRobotics EngineeringAerial RoboticsRobot DynamicsRoboticsCollision AvoidanceDynamic ControlQuadrotor DynamicsRobot SafetyRobust ControlSafety RequirementsControl Barrier FunctionsControl Lyapunov FunctionsLyapunov ControlLyapunov FunctionsSafety ConstraintsSafety-critical ControlSequential Quadratic ProgrammingUnderactuated Dynamics
Aerial robots such as quadrotors are subject to range, power, communication, and collision constraints, which have to be strictly enforced for safe operation. Such constraints are dynamic constraints and require the state of the system to be forward-invariant with respect to a safe set. We propose a control design method for a planar quadrotor subject to collision safety constraints. The proposed controller explicitly considers the nonlinear underactuated dynamics of a planar quadrotor and formulates a sequential quadratic program that uses time-varying control Lyapunov functions to guarantee stability and time-varying control Barrier functions to guarantee safety. In particular, we introduce an augmented control Barrier function that enforces a safety region in position space while explicitly taking into account the quadrotor's orientation. We present several numerical results to validate the proposed controller, including (a) setpoint regulation subject to a single static obstacle, (b) trajectory tracking with multiple static obstacles, and (c) trajectory tracking with a dynamic obstacle.
25
Constrained model predictive control: Stability and optimality
D.Q. Mayne, James B. Rawlings, Christopher V. Rao et al. · Automatica · 2000 · 8.5K citations