Neural-Learning-Based Finite-Time Trajectory Tracking Control for Robotic Manipulator with Input Friction

Due to the unknown nonlinear friction, there was a barricade preventing the precision of manipulators from further improvement. To overcome this challenge, the dynamic model of 2-degrees of freedom robot manipulator based on LuGre friction is established in this paper. The adaptive sliding mode observer is used to estimate the immeasurable states, meanwhile the neural network to approximate the friction. On this foundation, a neural-learning-based finite-time trajectory tracking control is designed to improve robustness. In particular, the closed-loop system stability is investigated by the Lyapunov theorem and computed the finite convergence rate thereafter. Finally, simulation results show that the control scheme has a better control effect of the manipulators.