Grey Box Identification of Fractional-order System Models from Frequency Domain Data

Abstract

In the following paper, the grey box system identification method is applied to the problem of fractional-order model identification from available frequency domain data. Oftentimes, this is the only feasible approach in case of industrial system identification where the capacity to perform time domain experiments is limited due to associated expenses. In this work, we comment on the potential of this approach and investigate the model parameter estimation method based on frequency domain data. Then, the implementation of the method in FOMCON toolbox for MATLAB is described. Finally, the resulting application is validated on the basis of a relevant academic example.