Model based deformable object manipulation using linear robust output regulation

Abstract

This paper presents an approach to the indirect simultaneous positioning task of deformable objects based on robust linear output regulation methods. Indirect control requires maneuvering control points defined on a deformable body to desired locations by manipulating points located elsewhere on the object. The proposed control scheme uses a linearization of the deformable object dynamics into a state-space model for which the inputs are the forces applied to the manipulation points and the outputs are the positions of the control points. Then, the indirect simultaneous positioning task is treated as a classical robust linear output regulation problem. The proposed controller can compensate for deformable model nonlinearities and material uncertainties. The controller performance is illustrated through simulation and experimental results obtained on a planar deformable object model. The experiment was conducted using a robot controlled to apply desired end-effector forces on the planar object.