Neural network adaptive position tracking control of underactuated autonomous surface vehicle

The present study investigates the position tracking control of the underactuated autonomous surface vehicle, which is subjected to parameters uncertainties and external disturbances. In this regard, the backstepping method, neural network, dynamic surface control and the sliding mode method are employed to design an adaptive robust controller. Moreover, a Lyapunov synthesis is utilized to verify the stability of the closed-loop control system. Following innovations are highlighted in this study: (i) The derivatives of the virtual control signals are obtained through the dynamic surface control, which overcomes the computational complexities of the conventional backstepping method. (ii) The designed controller can be easily applied in practical applications with no requirement to employ the neural network and state predictors to obtain model parameters. (iii) The prediction errors are combined with position tracking errors to construct the neural network updating laws, which improves the adaptation and the tracking performance. The simulation results demonstrate the effectiveness of the proposed position tracking controller.