Transient Tracking Performance Guaranteed Global NN Control of Robot Manipulator

In this paper, a neural network (NN) tracking controller is developed for robot manipulators. To deal with the unknown dynamics of the robot systems, radial basis function neural network (RBFNN) is integrated into the adaptive control design to compensate for the uncertainties. In order to achieve the predefined transient tracking performance, time-varied barrier Lyapunov functions (BLFs) are synthesized for the control design. A switching mechanism is incorporated into the controller to achieve a global uniformly ultimately boundedness (GUUB) stability, extending the conventional semiglobal uniformly ultimately boundedness (SGUUB) stability of NN control systems, such that requirement on the initial conditions can be relaxed. Simulations studies are performed to illustrate the effectiveness and efficiency of the proposed control scheme using a robot manipulator model.