The global exploration and exploitation of the abundant resources in the deep sea has sparked interest in the study of dynamic positioning (DP) systems. This study focuses on the robust adaptive neural network (NN) tracking control design for DP of vessels in the presence of model uncertainties, environmental disturbances, input saturation, and output constraints. The guided DP tracking control problem is first addressed based on reference filters. Then, the NN technology and robust term are fused into vectorial backstepping procedure to realize the adaptability to uncertainties and the robustness against disturbances. To satisfy prescribed output constraint specifications on tracking errors, prescribed performance control (PPC) method is integrated into proposed control design. In addition, an auxiliary system is used to cope with actuator saturation, which can reduce the abrasion of the actuators. Finally, to verify the effectiveness of the proposed control design, all comparative studies are performed in a joint high-fidelity simulation framework, and the simulation results on a shuttle tanker illustrate the effectiveness and superiority of the proposed DP control strategy.
The DP control system can continuously activate the its propellers to resist the sea loads (wind, waves, currents, etc.), and automatically control the position/heading of vessel (Peng et al., 2021). With the rapid development of industry and the ever-increasing demand for marine resources, more and more vessels serving complex maritime segments are also equipped or retrofitted with DP capabilities, such as semi-submersibles for floating installations.
Broadly speaking, not only limited to station-keeping, DP control of low-speed maneuvering operations also cater to practical requirements (Sørensen, 2011). For large changes in the setpoint, the controller will execute sudden large commands with large errors, while the guidance system can compute the appropriate smooth reference signal for the vessel to follow subject to mission or geometric constraints, e.g., the desired path or trajectory, intended to circumvent undesired motion behavior of the vessel (Fossen, 2011). Therefore, the guidance-based DP tracking control for DP of vessels is more practical than the DP control directly.