Document Type : Original Article
Author
Shahrood University of Technology
Abstract
This paper presents a new robust model-free control strategy for uncertain flexible-joint robot manipulators subject to actuator saturation and external disturbances. The proposed framework integrates a backstepping-based adaptive control architecture with the Bernstein-Schurer-Stancu (BSS) operator as a universal approximator to compensate for unknown system dynamics and uncertainties without requiring precise mathematical modeling. To circumvent the complexity problem, a command-filtered control methodology with an error compensation mechanism is incorporated. Additionally, an auxiliary compensation system is developed to mitigate the adverse effects of input saturation, with its states directly utilized to construct the compensated error surfaces that drive the adaptation laws for the BSS operator coefficients. The uniformly ultimately bounded stability of the closed-loop system is rigorously established through recursive Lyapunov analysis. Comparative simulation studies on a flexible-joint robot manipulator demonstrate that the proposed BSS operator-based approach achieves superior tracking performance, faster convergence, and reduced computational overhead compared to conventional fuzzy control scheme. The theoretical results validate the effectiveness of the proposed methodology for satisfactory control of flexible-joint robots under practical constraints.
Keywords
- Model-free control
- Bernstein-Schurer-Stancu operator
- flexible-joint robot
- command-filtered approach
- input saturation
Main Subjects
)