Impedance control stability properties in common implementations

Various implementations of impedance control have been suggested, usually based on idealized models of the physical system. This work considers the nonideal, practical effects of computation and/or communication delays and manipulator dynamics on the behavior of two primary approaches to impedance control. The results are cast in the form of stability boundaries, i.e. the relationships between desired impedance parameters which cause marginally stable behavior in the overall system. These stability boundaries are compared for the two primary implementations, and relative benefits of each approach are discussed. These comparisons provide the basis for quantitative tradeoffs, allowing selection of control implementation approaches suited for particular manipulators or allowing quantitative decisions to be made in manipulator system design.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>