Robust control oriented model set validation with input perturbations in a stochastic framework

The problem of validating a model set described by a linear fractional transformation with input perturbations is discussed under a stochastic framework. An analytic expression has been derived for its point unfalsified probability, on the basis of a necessary and sufficient condition for a model set being consistent with frequency domain plant input-output data. Numerical simulation show that unfalsified probability is appropriate in quality evaluation for a linear fractional uncertain model set, even when experimental data is from a closed-loop system.