A Model-Free Fixed-Time Terminal Sliding Mode Control Method for Robot Systems

This paper addresses some of the key challenges associated with sliding mode control (SMC)–based trajectory tracking for robotic manipulators. The primary objective is to enhance the accuracy of the follow-up while minimizing chattering behavior and ensuring robustness against uncertainty components and disturbances. Furthermore, this paper aims to eliminate the reliance on complete or partial knowledge of system dynamics, especially in the context of model-based SMC methods. Finally, the paper focuses on achieving fixed-time global convergence rates for the system. To address these issues, we first design a fixed-time terminal sliding mode control (TSMC) method and analyze its limitations. We then propose a model-free fixed-time TSMC strategy to overcome these limitations, removing the necessity for a dynamic model of the robotic system, thus enhancing practical implementation. The proposed approach ensures the tracking error converges to zero within a fixed time, and its fixed-time global stability is rigorously demonstrated using Lyapunov theory. Extensive simulations on the SAMSUNG FARA AT2 robotic arm were conducted to validate the proposed strategy. The results showed improved tracking performance and a significant reduction in chattering in the control input.