This paper investigates the problem of ship course control in the presence of model uncertainties, external disturbances, and actuator saturation. A high-performance autopilot is developed based on a direct neural network adaptive dynamic surface control (DSC) framework integrated with deep reinforcement learning. To compensate for lumped uncertainties arising from unmodeled dynamics and disturbances, a radial basis function (RBF) neural network is employed to provide online approximation within the control design. Moreover, the actuator saturation constraint is explicitly incorporated into the controller, avoiding performance degradation commonly encountered in conventional DSC schemes.To alleviate the reliance on manual parameter tuning, the controller parameter adaptation is formulated as a continuous-action optimization problem and solved using a deep deterministic policy gradient (DDPG) algorithm. The DDPG agent learns an optimal tuning policy by maximizing a reward function that penalizes course tracking errors, excessive control variations, and energy consumption. Simulation results demonstrate that the proposed method achieves improved tracking accuracy, smoother control inputs, and enhanced robustness under complex operating conditions, thereby validating the effectiveness of the DDPG-based adaptive tuning strategy for autonomous ship navigation.

To address the issue of low heading tracking efficiency caused by nonlinear dynamic characteristics in ship heading motion, this paper proposes a neural network-based adaptive hyperbolic tangent control method for ship heading. By designing a second-order system robust controller, a saturated auxiliary design system is introduced into the regulator for direct internal compensation, enhancing the system's anti-interference capability under complex operating conditions. Meanwhile, hyperbolic tangent nonlinear modification is incorporated into the control strategy to optimize the output characteristics of control signals. The controller adopts a backstepping approach to design virtual control laws for trajectory tracking and utilizes the Radial Basis Function (RBF) of neural networks to approximate the uncertain parts of the ship model. The control algorithm is simulated and tested in the MATLAB environment, and its tracking effect is analyzed. Simulation results show that the control algorithm can ensure the stability of the closed-loop system under conditions of dynamic changes in system parameters, external disturbances, and uncertainties, and effectively solve the nonlinear problems in ship traffic control during trajectory tracking. The controller is designed concisely, meets the requirements of engineering practice, improves ship maneuverability, and has reference value for ship control.