Fixed-time neural network trajectory tracking control for underactuated surface vessels

This paper provides a robust fixed-time trajectory tracking controller for underactuated surface vessels (USVs) suffering from unmodeled dynamics and external disturbances. Initially, a novel model transformation is firstly applied for the possible application of sliding mode control technology. Then, fixed-time convergence for tracking errors can be guaranteed by stabilizing the transformed system. During the design process, a constructive sliding mode surface is structured with the application of hyperbolic tangent function, which could ensure the settling time of the designed system independent of initial states. To accommodate unknown system dynamics and perturbations, the Minimum-Learning-Parameter based neural network and adaptive updating laws are adopted. Theoretical analysis shows that tracking errors enjoy practical fixed-time stability under the proposed controller. Numerical simulation results illustrate the effectiveness and superiority of the proposed control scheme.

Automated summary

This paper presents a robust fixed-time trajectory tracking controller for underactuated surface vessels, which can handle unmodeled dynamics and external disturbances. The design involves the use of sliding mode control technology and neural networks, with theoretical analysis and numerical simulation results demonstrating the effectiveness and superiority of the control scheme.