A novel adaptive neural network-based time-delayed estimation control for nonlinear systems subject to disturbances and unknown dynamics

• Published in: ISA transactions Vol. 142; pp. 214 - 227
• Main Authors: Truong, Hoai Vu Anh, Nguyen, Manh Hung, Tran, Duc Thien, Ahn, Kyoung Kwan
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.11.2023
• Subjects: Backstepping control; High-order systems; Radial basis function neural network (RBFNN); Time-delayed estimation (TDE); Backstepping control; High-order systems; Radial basis function neural network (RBFNN); Time-delayed estimation (TDE)
• ISSN: 0019-0578; 1879-2022; 1879-2022
• DOI: 10.1016/j.isatra.2023.07.032

This paper presents an adaptive backstepping-based model-free control (BSMFC) for general high-order nonlinear systems (HNSs) subject to disturbances and unstructured uncertainties to enhance the system tracking performance. The proposed methodology is constructed based on the backstepping control (BSC) with radial basis function neural network (RBFNN) -based time-delayed estimation (TDE) to overcome the obstacle of unknown system dynamics. Additionally, a command-filtered (CF) approach is involved to address the complexity explosion of the BSC design. As the errors arising from approximation, new control laws are established to reduce the effects in this regard. The stability of the closed-loop system is guaranteed through the Lyapunov theorem and the superiority of the proposed methodology is confirmed through a comparative simulation with other model-free approaches. •Nonlinear systems with completely unknown dynamics and unstructured uncertainties.•An RBFNN-based TDE to reduce a number of estimated unknown terms.•Disturbance observer to mitigate effects from redundant process errors.