Fractal system identification for robust control - the moment approach

In the last decade, several authors focus on fractional-order systems (called also fractal systems), which can describe reality better than classical integer-order models. The fractional-order description is based on generalized fractional order derivative and integral operators. This paper deals with fractional order system identification for the purpose of a robust controller design. The proposed method combines two pieces of information: a priori information and experimental data. Firstly, the class of all admissible process models is defined as the set of all a priori admissible transfer functions which are consistent with experimental data (moments of the system). Using a value set approach, the parameterization of all so called extreme transfer functions (those that are important for robust design) is presented.