This paper considers the application of the discrete-time
structured singular value to assess the robustness of systems subject to real parametric
uncertainties. Since the associated uncertainty set is strictly real, the
resulting mu bounds may be discontinuous irrespective of the fineness of the
frequency sweep. It is therefore necessary to seek other more reliable methods
of test. One such method detailed is a discrete robust stability state-space
test that is not dependent on a frequency sweep. It is shown that this is a
discrete skewed structured singular value problem. For illustration purposes, a
robustness analysis using both a frequency sweep and state-space transformation
is performed on a forward converter with a real parameter uncertainty. The
results are detailed and critically assessed.
History
Publication
Irish Signals and Systems Conference (ISSC); pp. 90-95