Polishing robot for PET bottle molds using a learning-based hybrid position/force controller

In this paper, a high precision polishing robot with a learning-based hybrid position/force controller is proposed for polishing PET (poly ethylene terephthalate) bottle molds. In polishing, the position control system loosely interferes with the force control system in suitable directions. The shape of a mounted abrasive tool, attached to the tip of a robot arm, is a ball-end type. When a PET bottle mold with curved surface is polished, not only the orientation of the mounted abrasive tool is fixed but also its revolution is locked. The motion of the mounted abrasive tool is feedforwardly controlled based on an initial trajectory calculated in advance. The trajectory is generated from cutter location data constituted from a CAM system. The trajectory is modified through the actual polishing processes so that the total force error in polishing becomes smaller. The surface is polished by a polishing force acting between the mold and the abrasive tool. The polishing force is assumed to be considered as a composite force of the contact and kinetic friction forces, in which the friction consists of Coulomb and viscous frictions. A few polishing experiments are conducted to show that the proposed system is effective for obtaining a polishing surface as achieved by skilled workers.