Minimal learning parameters-based simplified robust adaptive saturated control for dynamic positioning ship with input magnitude and rate saturations

In this study, a simplified robust adaptive saturated control strategy based on minimal learning parameters is developed for the dynamic positioning of ship with input magnitude and rate saturations, unknown external environment disturbance and dynamic uncertainties. Firstly, an augment model is developed to restrict the boundedness of the actuators while the input magnitude and rate saturations can be shown in a high-order model. Then, radial basis function neural networks are applied to formulate a new control law via the velocities backstepping method to hand with the dynamic uncertainties. In particular, the minimal learning parameters method is used to reduce the computational complexity, with a single parameter needing to be updated in each step of backstepping. Meanwhile, robust adaptive compensation terms are introduced into the design of virtual and actual control while the error caused by neural networks is mitigated. In line with the Lyapunov theory, the uniformly ultimately boundedness of all signal in closed-loop control system is demonstrated. Finally, simulation is performed to illustrate the advantage and effectiveness of the proposed control strategy.