Hybrid Model Predictive Control For Crosswind Stabilization Of Hybrid Airships

A hybrid airship is an aerial vehicle that generates lift by leveraging both buoyancy and aerodynamic principles. The operation of such a vehicle can be limited by its high susceptibility to crosswinds during taxiing, take-off and landing. With the goal to mitigate this issue, this paper proposes a novel controller design for a stabilization system consisting of wing tip thrusters. Due to the response of the vehicle to wind disturbances (e.g. lifting off a wheel during taxiing), modeling it as a hybrid dynamical system is appropriate. A novel, customized hybrid model predictive control (MPC) scheme is proposed for crosswind stabilization. As shown in simulation as well as in experimental results in controlled and realistic environments, the proposed control scheme succeeds in stabilizing the vehicle despite artificial or actual wind disturbances, even in scenarios where simple linear MPC fails. Simultaneously, our approach is computationally efficient enough to run on an onboard computer.