Interactive guidance system for kinematically-redundant manipulator

摘要

Starting with a controlled end-effector target movement and the actual value (qi) of the manipulator joint position, a new joint position (qi+i) is calculated, taking account of a merit function (f(q)) parametered by non-negative weighting values (aj,bj), also of path restrictions by physical joint stops (qmin,qmax) for maximum joint speed, and of the kinematic equation represented by the Jacobi-Matrix. The new joint position determines the new values for the joint-regulator. The merit function is the sum of the energy and reference position criteria, the respective formulae for which are (q-qi) tdiag (aj) (q-qi) and (q-qref) tdiag (Bj) (q-qref). The value (qref) is a predetermined joint position so determined that the sequence of the calculated joint-position values runs closely round the reference-position value. Using a joint-position value (qo) as a starting point, a permissible optimisation vector is determined relating to all active auxiliary conditions which indicate which path limitations are reached.