Model-free Fault-tolerant Control of Robotic Manipulator Subject to Actuator Fault: A Robust Control Approach Based on Function Approximation Technique with Considering its Accuracy

Author

Department of electrical engineering, Shahrood University of Technology

10.22044/jsfm.2026.17573.4053

Abstract

This paper introduces a novel robust fault-tolerant control (FTC) framework for electrically-driven robot manipulators subject to simultaneous actuator faults and time-varying external disturbances. To address the inherent uncertainties in both the manipulator and motor dynamics, a model-free approach is developed using function approximation technique (FAT) as universal approximators, eliminating dependence on precise system modeling. A core contribution is the formulation and incorporation of a approximation quality index into the adaptive update laws in the framework of FTC. This error-driven refinement mechanism, which quantifies the instantaneous fidelity of the identifier, markedly accelerates parameter convergence and enhances the accuracy of online uncertainty estimation. The learning-based design is further fortified to actively estimate and compensate for lumped perturbations, including fault-induced effects and exogenous disturbances. Within a voltage-based control structure, the proposed scheme ensures that all closed-loop signals remain uniformly ultimately bounded, as rigorously proven through Lyapunov stability analysis. Comprehensive simulations on a two-link robotic manipulator under fault effects demonstrate the superior performance of the proposed controller in maintaining trajectory tracking, uncertainty estimation and disturbance attenuation compared to existing related methods.

Keywords

Main Subjects

Journal of Solid and Fluid Mechanics
Volume 16, Issue 3
July and August 2026
Pages 233-247

Files

Share

How to cite

Statistics