Research Of Kinematics Singularity And First-Order Redundant Kinematic Planning Of Robotic Manipulators

Kinematics and kinematic planning play a significant role in robotics. They are inevitably required by almost all robotic manipulators which are applied to industry, space detection, and also intelligence which has been increasingly thought of by researchers.In this paper, damped least-square method utilized to overcome kinematic singularity of robotic manipulators was improved, which achieved more accurate tracking compared to the traditional method. Singular value decomposition (SVD), to which was attached great importance in singularity-robust inverse kinematics and kinematic control of redundant manipulators, was investigated.Also, an overwhelmingly utilized algorithm of SVD was improved and some possible problems encountered in joint failure issues and redundant issues when using this algorithm was deeply discussed. First-order kinematic planning scheme was described in detail so as to lay a solid foundation for trajectory planning.Referring to kinematic control of redundant manipulators, extended-Jacobian scheme and gradient-projection method of optimization scheme was mainly discussed. Extended-Jacobian scheme achieved Task-priority that main tasks were preferably guaranteed while secondary tasks were accomplished through self-motions of the robotic manipulator. And gradient-projection method focused on changing robotic motion from an optimizing perspective, which achieved some performance optimizations while fulfilling required tasks.With regard to Cartesian trajectory planning, we investigated a first-order scheme, which relied on direct Cartesian trajectory planning. The joint velocity could be guaranteed smoothly continuous so long as the end velocity of the manipulator was smoothly continuous, and when joint velocity limit was added, the joint velocity could be guaranteed continuous.