Nonlinear attitude control design and verification for a safe flight of a small-scale unmanned helicopter

Autonomous small unmanned helicopter systems have been widely studied in the last decades. These systems are extremely agile due to their energy efficiency, overall costs and high levels of maneuverability compared to manned helicopters. This allows them to be used in urban environments for different applications such as search and rescue, aerial stunts for movie industry, fire fighting, surveillance, etc. Such applications require the control system to be robust and safe since a fault may lead to environmental damage and endangering human life. For reasons of the very high safety requirements, in this paper we propose a robust control design and also introduce formal verification of control for small-scale unmanned helicopters. The controller proposed is based on dynamic inversion control for a 3-DOF (degree-of-freedom) attitude dynamics while taking into account the system modelling uncertainty with variable payloads and external disturbances of wind. An invariant set called control-enabled-set is defined for flight envelope, which represents the dynamical state vectors comprised of the attitude and rotation rates, for which the stable control of the craft is feasible with our control scheme. Then the controller is verified using formal methods represented by MetiTarski automated theorem prover to ensure controller stability and robustness. Our approach also paves the way to the possibility that the autopilot system monitors whether it is getting near the boundary of its flight envelope, in which case it can propose or plan and execute an emergency landing to a safe location.