New controller design for a robot manipulator

This paper presents a new systematic method to design a controller using the quantitative feedback theory (QFT) for a robot identified as a non-minimum phase model. The robot manipulators are highly nonlinear with an uncertain environment. A linearised transfer function of a laboratory robot is developed using the off-line system identification. The effects of nonlinearities are accounted by describing the linearised model parameters as structured uncertainty. The QFT design procedure is carried out to design a robust controller that satisfies performance specifications for tracking. The designed fixed-gain controller is easy to implement.