Single-parameter-learning-based robust adaptive control of dynamic positioning ships considering thruster system dynamics in the input saturation state

The focus of this paper is presented on robust adaptive dynamic positioning control in the face of thruster system dynamics. In the maritime domain, it considers the issues of model parameter ingestion, unknown time-varying disturbances, and input saturation. First, a finite-time convergent disturbance observer is used for the online estimation of unknown time-variant disturbances. Additionally, the model ingestion problem is also solved with a single-parameter learning neural network. Furthermore, a robust control term is introduced to account for undesired errors. Then, the thruster dynamics equation is considered to solve the issue of thruster dynamics characteristics in the designed process of the controller. Finally, the input saturation problem is addressed with a finite-time auxiliary dynamic system. The suggested dynamic positioning control approach allows the ship to retain the required position and direction, as demonstrated. Respectively, all control variables in the dynamic positioning control system are consistent and ultimately bounded. At last, the proposed dynamic positioning control method was validated through the experimental simulations on the supply vessel Northern Clipper.