Hierarchical Predictive Sliding Mode Control for Underactuated UAV: An FAS Approach

Abstract

This study introduces a hierarchical predictive integral sliding mode control (HP-ISMC) methodology for the quadrotor uncrewedaerial vehicle (UAV) based on the fully actuated system (FAS) approach. The proposed framework adopts a hierarchical control scheme, structured into three core components: system remodeling, controller design, and experimental validation. Unlike existing methods that directly linearize the system without fully accounting for its inherent nonlinearities and underactuation, we propose a hierarchical three-loop controller that explicitly tackles the complexities of nonlinear underactuated quadrotor UAV tracking control, particularly in the presence of disturbances and state constraints. Our approach reformulates the quadrotor UAV model into a multiple-input multiple-output (MIMO) sub-FAS, establishing a solid foundation for subsequent controller development. The HP-ISMC law is then designed within these relatively independent three loops. The first loop eliminates nonlinearities and establishes a decoupled linear system with disturbances and constraints, thereby simplifying the handling of uncertainties and constraints. The second loop then employs ISMC to fully compensate for the matched uncertainties in the linear systems obtained from the first loop. The third loop enforces the tightened state constraints and guarantees system stability while addressing potential unmatched uncertainties through predictive control. Finally, simulations and experiments are conducted to verify the effectiveness of the proposed method.