A new tuning procedure for PID control of rigid robots

This paper is concerned with classical PID control of rigid robots. We introduce a tuning procedure for selection of the PID gains ensuring asymptotic stability in a domain which can be enlarged arbitrarily. The novelty of our approach relies on the fact that conditions for stability are formulated as expressions that have to be satisfied at each joint instead of conditions on norms of gain and parameter matrices as reported previously. This allows better performances than those obtained using tuning procedures previously reported in the literature. (C) Koninklijke Brill NV, Leiden and The Robotics Society of Japan, 2008.