Chattering-free neural sliding mode control of robotic manipulator with actuator dynamics

In this paper, a chattering-free neural sliding mode control (CFNSMC) scheme is proposed for robotic manipulators with considering actuator dynamics. A full-order sliding mode surface is constructed to develop the finite-time controller such that the chattering and singularity problems can lie avoided since the derivatives of terms with fraction powers do not appear in the actual control law. Besides a radial basis function neural network (RBFNN) is employed to approximate the system uncertainties and nonlinearities. With the proposed scheme, no singularity and chattering problems exist in the controller design, and the convergence of the position tracking error is guaranteed via the Lyapunov synthesis. Comparison simulations of a two-link robotic manipulator are provided to verify the effectiveness of the proposed method.