H∞ finite-time prescribed performance trajectory tracking control of MSVs with self-adjusting neural networks

The paper proposes a robust adaptive finite-time prescribed performance trajectory tracking control scheme for marine surface vessels (MSVs) under unknown external disturbance and model uncertainty, based on H∞ control and self-adjusting neural networks (SANN). Firstly, to tackle the challenge of unknown velocity, a high-gain observer is introduced. Secondly, SANN is designed to tackle model uncertainty, achieving a balance between the optimal number of neurons and the best expected performance, thus saving network resources. Simultaneously, finite-time (FT) auxiliary functions, error transformation functions, and H functions are introduced, ensuring a bounded expected decay level with L2 norm within FT. Finally, by combining backstepping, SANN, prescribed performance control and FTH∞ control methods, the system ensures the achievement of practically finite-time stable using fewer network resources under prescribed performance conditions. Simulation results validate the effectiveness of the proposed controller. •A finite-time prescribed performance H∞ trajectory tracking control method for surface vessels is proposed.•Antidisturbance problems of vessel control systems subject to unknown velocities are discussed.•An output feedback control structure based on self-regulating neural network is constructed.