SUMMARY

Autorotation maneuvers are performed by pilots to safely land after an engine failure in a helicopter. These maneuvers are difficult due to the small window for successful timing, the wide range of possible entry conditions, and the potentially catastrophic consequences of a mistake. This work focuses on automation of the autorotation maneuver and the development of cues to aid in piloted autorotations. Specifically, this work is broken into three main categories: landing point tracking and reachability determination near landing, model predictive control through touchdown, and a landing site reachability determination algorithm and pilot cue. The landing point tracking scheme utilizes a biomimetic strategy called Tau-Theory to generate sub-optimal trajectories nearly instantaneously. A point-mass physical model of the helicopter is then applied to predict states and control along an input trajectory. These predicted states can be used to determine the feasibility of the given trajectory. A set of candidate trajectories can be generated and evaluated using these methods to find a sub-optimal set of reachable landing points. This set of landing points can be used to cue a pilot to aid with landing point selection. The Nonlinear Model Predictive Control (NMPC) method proposed here offers several potential benefits over existing control methods including the capability to intelligently balance state constraints on three outputs using only two control inputs. This multi-input multi-output NMPC method solves for the optimal control inputs in closed form using the same point-mass helicopter model employed above. This ensures deterministic runtime and enables real-time execution, a necessities for aerial vehicles. An algorithm to determine the reachable landing points in descent phase is presented along with implementation of two Head Up Displays driven by the algorithm. Limited piloted studies are presented to evaluate the usefulness of such a cue.