Fuzzy Finite-Time Sliding Mode Control of Euler-Lagrange Systems with State/Error Constraints

This paper studies the fuzzy finite-time nonsingular terminal sliding mode (FTNTSM) control problem for state/error-constrained and uncertain Euler-Lagrange systems with several contributions: (1) a modified FTNTSM manifold is designed to ensure finite-time stabilization of error dynamics and prescribed tracking accuracy; (2) a new barrier Lyapunov function (BLF) candidate, not only intended for sliding mode control design, but also devoted to constraint satisfaction, is constructed. The BLF candidate is a universal one, as it works uniformly for both constrained and unconstrained systems without changing the control structure; and (3) a fuzzy logic system is employed to dynamically compensate modeling inaccuracies and external disturbances. Meanwhile, a terminal dynamics-based gradient descent algorithm with an accelerated learning process is established to train the weight online. Numerical simulations on a robot manipulator demonstrate the theoretical findings.