Towards a state-space polytopic uncertainty description using subspace model identification techniques

We introduce a method for the estimation of the uncertainty of a model, identified using subspace identification methods. The key in this method is the calculation of the first order approximation of the relation between the perturbation of the data and the error of the elements in the identified state-space matrices. Using first order approximation we can find a polytopic description of the uncertainty region of the identified model. This polytopic description can for instance be used to develop a robust(ified) controller for the plant