Identification for Control Package ( version 1 . 1 b 4 , February 2001 ) – A Guided Tour –

This software package provides algorithms for control oriented identification, based on measured data. The techniques implemented in particular are non-stationary stochastic embedding, model error modeling using prediction error methods and set membership estimation. Analysis of the obtained model sets is possible using ν-gap methods. A robust controller design with respect to parametric and/or dynamical errors is implemented as well. Most of these methods rely heavily on the usage of orthonormal basis functions, which are described in an object oriented fashion in this package.