Robust adaptive control for dynamic positioning ships in the presence of input constraints

This note investigates the dynamic positioning control problem of full-actuated marine vehicles in the presence of system uncertainties and input constraints. In the control algorithm, the ship model uncertainty is approximated by the radial-basis-function neural networks. The effect of input saturation is analyzed by the auxiliary system, states of the auxiliary system are employed to develop the control scheme. By virtue of the dynamic surface control technique, the inherent problem of “explosion of complexity” problem occurred during conventional Backstepping design framework is avoided. The semi-global uniformly ultimately bounded stability of the closed-loop is guaranteed by Lyapunov theory. Since the minimal learning parameterization based adaptive law is derived, the number of parameters updated online is reduced to 6. Consider the servo-system uncertainty, the control inputs of interest are achieved as the measurable propeller pitch whose characteristics would facilitate the implementation of the algorithm in the practical engineering. Finally, by employing a supply vessel as the plant, comparison simulations are conducted to demonstrate the effectiveness and robustness of the proposed algorithm.