A local approach for the modeling of linear parameter-varying systems based on transfer function interpolation with scaling coefficients

In this paper, we propose a novel modeling method for linear parameter-varying systems. First, a set of frequency-domain data samples of the system transfer function is computed for fixed values of the scheduling parameters. Then, a system identification step is performed to generate a set of rational models called root macromodels. Finally, a linear parameter-varying model is built by means of interpolation of the root macro-models at the transfer function level with scaling coefficients. This interpolation scheme is different from standard interpolation of state-space matrices for the modeling of linear parameter-varying systems. A measurement-based example is used to validate the proposed linear parameter-varying modeling technique.