Generation and Evaluation of the Workspace of a Manipulator

Abstract

An algorithm is presentedfor noniterative generation of the workspace of a general n-R manipulator. It is shown that if special techniques are used for the reduction of points, such algorithms, which otherwise require large storage, become computationally feasible. For the evaluation of the workspace, approach angles and approach lengths for points in the workspace are analyzed by using a stable iterative algorithm for inverse kinematic analysis. One modification makes it a strict descent algorithm with respect to certain measures of deviation from the target position; another uses a variable convergence criterion to improve its speed and accuracy in the determination of the limiting configurations.