Search

In order to tackle the marine practical constraints, for example the actuator faults, the dead‐zone input, an improved composite adaptive neural control algorithm is proposed for dynamic positioning vehicles in presence of the unknown external disturbances. In the algorithm, the robust neural damping technique is employed to remodel the system model uncertainty and suppress the external interference. As for the dead‐zone input, the dead‐zone inverse model is constructed to derive the corresponding compensating terms. That could effectively release the constraints from the actuator faults and the dead‐zone non‐linearity. Furthermore, for merits of the composite intelligent learning method, one designs the serial‐parallel estimation model to estimate the related velocity variables. The corresponding prediction error could be applied in the design of adaptive law. That could effectively improve the accuracy of parameter estimation and facilitate the robustness of the closed‐loop system. The semi‐global uniformly ultimately bounded stability is guaranteed for all error signals in the closed‐loop system by utilizing the Lyapunov theory. Finally, the validity of the proposed algorithm is demonstrated through the simulation experiments. [ABSTRACT FROM AUTHOR]