A study of neural network control of robot manipulators

The basic robot control technique is the model based computed-torque control which is known to suuer performance degradation due to model uncertainties. Adding a neural network (NN) controller in the control system is one eeective way to compensate for the ill eeects of these uncertainties. In this paper a systematic study of NN controller for a robot manipulator under a uniied computed-torque control frame work is presented. Both feedforward and feedback NN control schemes are studied and compared using a common back-propagation training algorithm. EEects on system performance for diierent choices of NN input types, hidden neurons, weight update rates, and initial weight values are also investigated. Extensive simulation studies for trajectory tracking are carried out and compared with other established robot control schemes.