Nonlinear Model Predictive Control on SE(3) for Quadrotor Trajectory Tracking and Obstacle Avoidance

Some recent contributions have emerged designing Nonlinear Model Predictive Control (NMPC) for UAVs. However, these approaches often split the problem into upper and inner layers, or attitude and position control, separately, which can be undesirable in complex tasks such as those involving optimal functional cost depending on position and attitude references simultaneously. Moreover, most of their controller's design does not handle the avoidance of representational singularities. Therefore, the present work proposes a NMPC formulated on the Special Euclidean group SE(3), which has a single optimization layer, for quadrotor safe trajectory tracking with obstacle avoidance capacity. A numerical experiment illustrates this proposal and is used to evaluate the controller's performance.