Robust control of robot arms including motor dynamics

A state model that incorporates robot arm dynamics together with joint actuator dynamics is formulated as a nominally linear uncertain system. The uncertainty is matched and it is shown to be bounded by a cubic expression of the norm of the joint errors. A nonlinear robust feedback controller is then derived which guarantees global ultimate boundedness of the actual system trajectories. Computer simulations are performed on a model of the PUMA 560 robot arm and the results have verified the feasibility of our controller with excellent trajectory tracking performance.