Documentation Center

  • Trial Software
  • Product Updates

complexify

Replace ureal atoms by summations of ureal and ucomplex (or ultidyn) atoms

Syntax

MC = complexify(M,alpha)
MC = complexify(M,alpha,'ultidyn')

Description

The command complexify replaces ureal atoms with sums of ureal and ucomplex atoms using usubs. Optionally, the sum can consist of a ureal and ultidyn atom.

complexify is used to improve the conditioning of robust stability calculations (robuststab) for situations when there are predominantly ureal uncertain elements.

MC = complexify(M,alpha) results in each ureal atom in MC having the same Name and NominalValue as the corresponding ureal atom in M. If Range is the range of one ureal atom from M, then the range of the corresponding ureal atom in MC is

[Range(1)+alpha*diff(Range)/2 Range(2)-alpha*diff(Range)/2]

The net effect is that the same real range is covered with a real and complex uncertainty. The real parameter range is reduced by equal amounts at each end, and alpha represents (in a relative sense) the reduction in the total range. The ucomplex atom will add this reduction in range back into MC, but as a ball with real and imaginary parts.

The ucomplex atom has NominalValue of 0, and Radius equal to alpha*diff(Range). Its name is the name of the original ureal atom, appended with the characters '_cmpxfy'.

MC = complexify(M,alpha,'ultidyn') is the same, except that gain-bounded ultidyn atoms are used instead of ucomplex atoms. The ultidyn atom has its Bound equal to alpha*diff(Range).

Examples

expand all

Complexified Uncertain Parameter

To illustrate complexification, create a uncertain real parameter, cast it to an uncertain matrix, and apply a 10% complexification.

a = umat(ureal('a',2.25,'Range',[1.5 3]));
b = complexify(a,.1);
as = usample(a,200);
bs = usample(b,4000);

Make a scatter plot of the values that the complexified matrix (scalar) can take, as well as the values of the original uncertain real parameter.

plot(real(bs(:)),imag(bs(:)),'.',real(as(:)),imag(as(:)),'r.')
axis([1 3.5 -0.2 0.2])

See Also

|

Was this topic helpful?