Extended balanced truncation for parameter-varying multidimensional systems

In this work, a novel technique of balanced truncation for model reduction of parameter-varying multidimensional systems in “Roesser Form” is proposed. The multidimensional system representation requires two operators defined on l 2 -space. The reduction is based on extended Gramians. Both, standard and improved error bounds for the model reduction problem are presented. The approach is demonstrated with the application to an illustrative example.