Adaptive fuzzy iterative learning control of robotic systems with time-delay outputs and input dead zone via observer

Abstract

This paper presents an adaptive fuzzy iterative learning control method for the output tracking problem of robotic systems with unknown time delay output and input dead-zone. A state observer is designed to estimate unmeasurable velocity variables. By introducing boundary layer function, the identical initial condition for most iterative learning control schemes is relaxed. The problem of possible singularity is overcome by using hyperbolic tangent function. By combining appropriate Lyapunov-Krasovskii functional and fuzzy logic systems approximation technique, the proposed control scheme can guarantee that the output tracking converges to the desired reference trajectory within an error tolerance and all the closed-loop signals remain bounded. Finally, simulation results demonstrate the validity of the proposed approach.