Filter
Top Products
A. TRANSLATION BETWEEN MATLAB µ-TOOLS AND Xµ 403
The advantage of this is that in order to restart, or reproduce, an iteration, one need
only save the previous controller. The µ-Tools approach requires saving the rational
approximation to the previous D-scales. The controller is a more applicable data object
to save and the saving of the previous D-scales depends on the quality of the rational
approximation. The disadvantage is that the upper bound in the next µ calculation
takes slightly longer as the closed-loop system does not have the numerical benefit of the
effects of the previous D-scales. The speed difference is likely to be insignificant in
practical problems.
Fitting D Scales
Rational transfer function fitting of magnitude data is required for the D-K iteration
and has been mentioned above. The relevant functions are summarized here.
Description µ-Tools function Xmath/Xµ equivalent
fit D scale data musynfit musynfit
phase calculation genphase mkphase
transfer function fitting fitsys fitsys, tfid
Note that the µ-Tools version of musynfit does not use the previous D scale magnitude
data. This has implications in the D-K iteration and is discussed in more detail above.
The alternative µ-Tools linear programming approach for D scale fitting, musynflp,is
not supported in Xµ. Similarly, the underlying linear programming approach to transfer
function fitting (µ-Tools functions fitmaglp, magdata) is not available in Xµ. The batch
functionality of the µ-Tools function, muftbtch, is available in the Xµ version of
musynfit.