Adaptive control of redundant manipulators via pseudovariables

An approach to resolving redundancy and controlling redundant manipulators in a unified manner is proposed. The redundancy is resolved by imposed additional constraints on the redundant arm's motion. The constraints may be physical constraints, or they can arise from necessary conditions for an optimization problem. The dynamical model of the redundant manipulator is described in a reduced-order space. A control architecture is specified based on the reduced-order model. An adaptive controller is proposed to account for inaccurate parameters in the dynamical model. A study on the global stability of the algorithm is presented.<
>