Simple control strategy of redundant manipulators by disturbance observer

This paper proposes a simple and high performance control method of redundant robot manipulators. Disturbance observers in joint space and task space are used to linearize the redundant robot manipulator in joint space and task space, respectively. The disturbance observers which estimate step disturbance and ramp disturbance are tested. A combination of the ramp disturbance observer in joint space and the step disturbance observer in task space gives the beet performance for the linearization of redundant robot manipulators. Moreover, computation of the inverse kinematics can be simplified drastically; i.e. the inertia matrix in task space and the null space term. The computation error due to the simplification is compensated by the disturbance observer in task space.