Equivalence of position-level and velocity-level redundancy-resolution schemes

After widely investigating the velocity-level redundancy resolution schemes for redundant robot manipulators, we focus one of our research directions on the relationship between different-level schemes, e.g., between position-level and velocitylevel redundancy-resolution schemes. In this paper, we propose a novel viewpoint; that is, there exists an equivalence between some position-level and velocity-level redundancy-resolution schemes. To demonstrate such a viewpoint, two theoretical proofs, based on the gradient-descent method and Zhang et al´s neural-dynamic method, are presented in this paper. Moreover, such two different-level schemes are solved by employing two different methods, i.e., MATLAB routine “FMINCON” and linear-variational-inequality (LVI) based primal dual neural network (in short, LVI-PDNN). Computer-simulation results based on a five-link planar robot arm not only verify the equivalence of position-level and velocitylevel redundancy-resolution schemes, but also demonstrate the superiority of the velocity-level redundancy-resolution scheme, as compared with the position-level redundancy-resolution scheme.