Neuro-adaptive Synchronization Control Design for Dynamically Positioned Marine Vessels with Unknown Input Gains

Dynamic positioning technology has become a critical support for offshore engineering operations with the increasing complexity. This study focuses on the neuro-adaptive syn-chronization control for dynamically positioned (DP) vessels with model uncertainties, sea loads, and unknown input gains. The synchronization control issue is reformulated as leader-follower configuration, with the objective of tracking the independent-controlled leader vessel in synchrony. Neural network (NN) techniques are applied in the backstepping control process to achieve the adaptive ability to the lumped uncertain dynamics. An improved adaptive control scheme driven by estimation errors is further introduced into the proposed finite-time neuro-adaptive synchronization control strategy. Given such a dynamic positioning control design, the synchronous tracking errors, adaptive control gains, and NN estimated weights can converge simultaneously within a finite time. In veiw of the proposed control scheme, it is theoretically guaranteed that the closed-loop system will remain semi-global practical finite-time stable (SGPFS). Finally, taking a shuttle tanker as an example, the simulation analysis verifies the designed DP control scheme.