An Adaptive EKF-FMPC for the Trajectory Tracking of UVMS

Abstract

This paper addresses the trajectory tracking problem for the constrained dynamical underwater vehicle-manipulator system (UVMS) in the presence of dynamic uncertainties, sensory measurement noises, and time-varying external disturbances. An adaptive robust fast computational optimal control scheme is presented for the large-scale multiple-input–multiple-output (MIMO) constrained dynamical UVMS. The structure of our proposed scheme is based on a fast incremental model predictive controller (MPC) embedded with an extended Kalman filter (EKF). The EKF estimation part is used to compensate the unmodeled uncertainties, sensory noises, and external disturbances. The fast MPC (FMPC) part is a trajectory-tracking controller designed by the terminal constrained MPC with an approximately fast computational cost function, which guarantees the optimal batch control process satisfying the state and input constraints. By resorting to a Lyapunov-based stability analysis, the proposed FMPC is proved to be asymptotically stable with a monotonic decrease in the cost function, and can guarantee the tracking performance of the constrained dynamical UVMS. Finally, the effectiveness of the proposed control scheme is verified through a series of comparative simulations of a six-degree-of-freedom (6DOF) vehicle–6DOF manipulator system.