Optimizing Computational Efficiency in Self-Motion Manifold Exploration for Robotic Manipulator

Abstract

This thesis presents an in-depth exploration of self-motion manifolds in robotic manipulators and the development and refinement of methodologies for efficient and effective path planning in high-dimensional task spaces. The research focuses on the intricate relationship between the degrees of freedom in robotic manipulators and the corresponding self-motion manifolds, particularly addressing the complexities that arise in higher-dimensional spaces. A significant contribution of this thesis is the development of novel methodologies that utilize graph theory and clustering algorithms for navigating through self-motion manifolds. These methods demonstrate marked improvements in path planning and optimization, outperforming existing techniques and building upon previous advancements in the field. The research also delves into the effects of joint limits on these manifolds, revealing how they influence the segmentation into distinct sub-manifolds and impact the robot’s movement capabilities.