Abstract: A backstepping control based dynamic positioning control algorithm is developed to address the dead zone in actuator characteristics as well as uncertainty in system model structure and parameters. Robust neural damping technique is used to appropriate model uncertainty and to construct inverse model of dead zone for compensating the dead zone characteristics of inputs. The system is proved Semi-Global Uniformly Ultimately Bounded by Lyapunov stability theory. Simulations are performed to demonstrate the effectiveness of the algorithm.