Disturbance Attenuation Predictive Optimal Control for Quad-Rotor Transporting Unknown Varying Payload

摘要

Quad-rotor is very suitable for payload transportation due to the merits of high maneuverability and free hovering. However, the unknown varying payloads can cause negative influences that act in forms of persistent disturbances and sudden changes, damaging flight performance especially the attitude stability seriously. Targeting the persistent disturbances, an entirely novel disturbance estimator (DE) which can estimate non-smooth disturbances in a highly accurate manner for feedback compensation is proposed in this paper. To deal with the sudden changes from prescribed references and the payloads that may induce too large overshoots and input surging, a type of predictive optimal controller, which considers tracking errors and their changing rates of a class of linear multiple-input-multiple-output systems, is developed. Simulation results show that the system enhanced by the DE has better control performance than the ones enhanced by the commonly used extended state observer or nonlinear disturbance observer. Compared with the typical control approaches, the proposed control scheme enables the quad-rotor attitude system more stable performance and more ideal inputs on both persistent disturbance and sudden change resisting during payload transportation.