Adaptive nonlinear H;∞ techniques applied to a robot manipulator

In this paper, three nonlinear H;_∞ control techniques used to control a robot manipulator are compared. The first technique consists in an explicit solution of the robotic H;_∞ control problem. It is found considering that the dynamic parameter matrices are exactly known. In the second, a linear parameterization is used to generate an adaptive control law in the presence of uncertain parameters. Finally, a neural network is considered when there is unmodeled dynamics. Results obtained from the experimental robot manipulator UArm II, using the three methodologies, are presented.