Transient Tracking Performance Guaranteed Neural Control of Robotic Manipulators with Finite-Time Learning Convergence

An adaptive finite-time (FT) neural control scheme is proposed for robotic manipulators, which could guarantee transient tracking performances in the presence of model uncertainties. With the introduction of an error transformation mechanism, the original constrained manipulator system can be transformed into an unrestricted system. Moreover, the FT neural learning algorithm motivated by the estimated weights error, under persistent excitation (PE) condition, can guarantee the estimated neural weights converge to a small neighborhoods around the optimal values in finite time. Subsequently, the adaptive FT neural controller could ensure uniformly ultimate boundedness of all the signals in the closed-loop system and guarantee prescribed tracking and neural learning performances. The simulation results are given to illustrate the feasibility of the algorithm and correctness of theoretical analysis.