H ∞ finite-time prescribed performance trajectory tracking control of MSVs with self-adjusting neural networks
Zhipeng Shen,Yi Nie,Zixuan Zheng,Sheng Dong,Haomiao Yu
DOI: https://doi.org/10.1016/j.oceaneng.2024.117016
IF: 5
2024-04-01
Ocean Engineering
Abstract: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 L 2 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.
engineering, civil, ocean, marine,oceanography