A bio-inspired approach for online trajectory generation of industrial robots

Abstract

This paper presents a bio-inspired approach for generating online rhythmic point-to-point motions commonly used by industrial robots. Rhythmic motions in biological systems are produced with neural circuits called central pattern generators (CPGs). A modified form of CPG network is designed and used to generate the rhythmic motion. A pick and place task is considered an industrial application of the proposed approach. Several pick and place applications are considered. It is demonstrated that with CPGs, online changes of rhythmic trajectories are successfully generated. The proposed approach is novel, enables online trajectory generation, improves machine life and yields excellent coordinated natural motion for the robotic systems.