Robust control of manipulators using Hamiltonian optimization

Although nonlinear H/sub /spl infin// control theory for systems with uncertainties has been developed, it has been rarely applied to robot manipulators. We derive Hamiltonian equations of motion and Hamiltonian matrix (Riccati equation) for robot manipulators to get an optimal controller. Using the H/sub /spl infin// framework, we specify the L/sub 2/ norm of the performance measure, which consists of the position and momentum error during the task when there are uncertainties in the dynamic model of robot manipulators. For practical application, a stationary solution of the Riccati equation for the end-point is suggested and evaluated by simulation.