Pose and Path Planning for Industrial Robot Surface Machining Based on Direction Fields

Abstract

This study proposes a pose and path planning method based on direction fields, addressing challenges in industrial robot surface machining. These challenges include limited machining accuracy and difficult path planning arising from the robot's low stiffness and high nonlinearity between joint and operational spaces. The method consists of two key components. Firstly, a smooth pose planning method is proposed, independent of specific machining paths. It guides robot pose by constructing the smooth 1-direction field on the freeform surface, ensuring smooth pose transitions between adjacent machining points. Secondly, a path planning method based on a 2-direction field is proposed. The approach generates a 2-direction field by considering the robot's performance and task requirements. After integrating the 2-direction field to obtain a scalar field, iso-lines are extracted as the machining paths. The results demonstrate that the proposed method can generate smooth machining poses and continuous paths, even for complex surfaces where other methods may fail to generate fully reachable machining paths. The proposed method could achieve different path performances, such as high manipulator-velocity-ratio (MVR), high tangential stiffness, and high machining efficiency, by tracking specific 2-direction fields.