Command filtered backstepping control of constrained flexible joint robotic manipulator


Here, an adaptive radial basis function (RBF) neural network (NN) backstepping controller is proposed for a class of input-constrained flexible joint robotic manipulators represented by strict-feedback form with unknown terms, external stochastic disturbance, and output disturbance. The proposed approach is robust against both deterministic and stochastic uncertainties and disturbances and copes with the control input amplitude saturation. Moreover, by deploying the minimal learning parameter method and command filter technique, the computational burden of derivative terms and adaptive terms greatly decreases. Considering the mean-value theorem assists us to avoid the need for having the input saturation bounds in prior. The suggested tracking control scheme mandates the closed-loop system states to be semi-globally bounded-in-probability. Also, a quartic Barrier Lyapunov function is utilized to force the tracking error to be confined within a pre-chosen small region around the origin. Eventually, a numerical simulation of a flexible joint robot manipulator with a single link is performed to show the effectiveness and performance of the developed control method.

Publication Title

IET Control Theory and Applications