High-Gain Observer-Based Line-of-Sight Guidance for Adaptive Neural Path Following Control of Underactuated Marine Surface Vessels

Abstract

This paper presents an adaptive neural network dynamic surface control law with time-varying sideslip compensation, which includes a geometric kinematic control design and a dynamic control design, for path following of underactuated marine surface vessels (MSVs) in the presence of unmeasured states, model uncertainties, unknown ocean environmental disturbances, and input saturation. First, a high gain observer based line-of-sight guidance law, capable of estimating and compensating the time-varying sideslip angle induced by the unknown time-varying ocean disturbances, is proposed to generate the desired heading for the dynamic control system of the MSV. Next, considering the input saturation, adaptive neural network dynamic surface velocity control laws, in which the model uncertainties and unknown ocean environmental disturbances of the system are compensated by adaptive radial basis function neural network and the problem of input saturation is solved by introducing auxiliary dynamic systems, are designed to realize the dynamic control target of the path following. The proposed control strategy of path following compensates the ocean environmental disturbances from the perspective of both the geometric kinematics and the dynamics, and all error signals of a closed-loop control system are proven to be uniformly ultimately bounded. The simulations validate the effectiveness of the proposed control strategy for path following of MSVs.