Jerk-level solutions to manipulator inverse kinematics with mathematical equivalence of operations discovered

摘要

By applying direct derivative dynamics (DDD or D3), gradient neural dynamics (GND) and Zhang neural dynamics (ZND) at joint-jerk level, three novel solutions (namely, of types Z1D1G1-D1Z1G1, D2Z1-D1Z1D1-Z1D2 and D1Z2-Z1D1Z1-Z2D1) are proposed, developed and investigated in this paper. It is the first time that a relatively complete hybrid solution framework is proposed at the joint-jerk level. Then, a mathematical equivalence of Z1D1 and D1Z1 operations is discovered. Besides, the path-tracking applications performed on a 5-link robot manipulator sufficiently illustrate the efficacy, accuracy and superiority of these solutions.