I. Introduction
Trajectory tracking for robot manipulators is one of the elementary objectives and is widely used in the robotic field [1]–[3]. In a realistic scenario, perfect knowledge of the dynamic model of the robot can never be assumed [4], [5]. Moreover, disturbance always exists and it may deteriorate the performance of the controlled systems significantly [6], [7]. Fast and high-precision tracking of robot manipulators is a topic that continues to challenge control theoreticians and engineers. Recently, finite-time stable control of robot manipulators has attracted much attention due mostly to its fast convergence and high steady-state accuracy [8]–[11].