REDUCED MOTION OF PARALLEL ROBOT MANIPULATORS DUE TO ACTIVE JOINT JAM

Abstract

In this study, the effect of active joint jam (lock) on the velocity of parallel manipulators is investigated. The reduced motion capability of parallel manipulators is discussed in light of the change occurring in the Jacobian matrix after active joint jam. A criteria equation is formulated to determine whether a given task velocity is achievable after an active joint jam. After active joint jam, the manipulator is either shut down, pending repair, or is used to complete the task after reducing the task requirements. This is done by classifying the task velocity space into a subspace in which the manipulator has to follow the desired task trajectory, and a subspace in which the manipulator could deviate from the task trajectory assuming this is permissible by the task (reducing task requirements). If the task is reduced such that the failed parallel manipulator becomes kinematically redundant relative to the reduced task requirements, optimization could be used to determine the post-failure end-effector trajectory. Optimum post-failure end-effector trajectories are discussed in terms of minimization of the end-effector velocity norm, minimization of the end-effector velocity error norm and minimization of the joint rates. The post-failure scenarios are implemented in a simulation of a 3-3 6-degree-of-freedom (6-DOF) Stewart-Gough manipulator.