Different-level simultaneous resolution of robot redundancy with end-effector path tracked and with joint velocity and acceleration both minimized

摘要

To remedy the phenomena of high joint-velocity/acceleration as well as nonzero end-motion joint velocity, a different-level bi-criteria minimization scheme is proposed for the redundancy resolution and end-effector path tracking of robot manipulators. The scheme combines the minimum two-norm joint-velocity and joint-acceleration solutions via two weighting factors. Physical constraints (e.g., joint-angle limits, joint-velocity limits, and joint-acceleration limits) are also incorporated simultaneously into the scheme formulation. More importantly, the proposed different-level bi-criteria scheme is finally formulated as one quadratic program (QP) to be solved at the joint-acceleration level. A simplified primal-dual neural network based on linear variational inequalities (LVI) is then employed for solving such a quadratic program and its original robotic scheme. Computer simulations based on a PA10-type manipulator and a three-link planar robot arm verify the efficacy, unification and flexibility of the proposed different-level bi-criteria minimization scheme.