Direct and composite iterative neural control for cooperative dynamic positioning of marine surface vessels

This paper considers the cooperative dynamic positioning of multiple marine surface vessels in the presence of dynamical uncertainty and time-varying ocean disturbances. The objective is to enable a group of vessels to automatically position themselves in a desired time-varying formation and track a reference position. By employing a dynamic surface control technique, distributed adaptive controllers are derived on the basis of the information of neighboring vessels. Neural network with iterative updating laws is used to accurately identify the dynamical uncertainty and time-varying ocean disturbances. Two types of adaptive laws are proposed and validated: (1) direct iterative updating laws based on the velocity tracking errors; (2) composite iterative updating laws based on the tracking errors and the prediction errors. For both cases, Lyapunov–Krasovskii functionals are employed to analyze the stability of the closed-loop network, and uniform ultimate boundedness of error signals are established. The key features of the proposed controllers are as follows. First, the information exchanges are reduced by employing a distributed control strategy. Second, by using the iterative updating laws, the mixed uncertainty including the internal model uncertainty and external time-varying ocean disturbances can be compensated. Besides, the proposed controllers are easier to implement in digital processors with the derivative-free updating laws. Third, the prediction errors and tracking errors are combined to construct the composite iterative neural control laws, which are able to achieve faster adaptation and improved performance. Comparison studies are given to show the effectiveness of the proposed methods.