Neural network output feedback control of robot manipulators

A robust neural network output feedback scheme is developed for the motion control of robot manipulators without measuring joint velocities, A neural network observer is presented to estimate the joint velocities. It is shown that all the signals in a closed-loop system composed of a robot, an obser ver, and a controller is uniformly ultimately bounded. This amounts to a se paration principle for the design of nonlinear dynamic trackers for robotic systems, The neural network weights in both the observer and the controlle r are tuned on-line, with no offline learning phase required. No exact know ledge of the robot dynamics is required so that the neural network controll er is model-free and so applicable to a class of nonlinear systems which ha ve a similar structure to robot manipulators. When compared with adaptive-t ype controllers, we do not require linearity in the unknown system paramete rs, or the tedious computation of a regression matrix, Simulation results o n 2-link robot manipulator are reported to show the performance of the prop osed output feedback control scheme.