Nonlinear Model Predictive Guidance for Fixed-wing UAVs Using Identified Control Augmented Dynamics

As off-the-shelf (OTS) autopilots become more widely available and\nuser-friendly and the drone market expands, safer, more efficient, and more\ncomplex motion planning and control will become necessary for fixed-wing aerial\nrobotic platforms. Considering typical low-level attitude stabilization\navailable on OTS flight controllers, this paper first develops an approach for\nmodeling and identification of the control augmented dynamics for a small\nfixed-wing Unmanned Aerial Vehicle (UAV). A high-level Nonlinear Model\nPredictive Controller (NMPC) is subsequently formulated for simultaneous\nairspeed stabilization, path following, and soft constraint handling, using the\nidentified model for horizon propagation. The approach is explored in several\nexemplary flight experiments including path following of helix and connected\nDubins Aircraft segments in high winds as well as a motor failure scenario. The\ncost function, insights on its weighting, and additional soft constraints used\nthroughout the experimentation are discussed.\n