Frequency Domain System Identification Toolbox for MATLAB: A Complex Application Example

Abstract Frequency domain system identification has been rapidly developing during the past years. The new comprehensive Matlab toolbox is a successful implementation of the methods of the whole identification procedure. Its core is a maximum likelihood estimate of parametric transfer function models, maybe with a delay, but also offers help for nonparametric transfer function identification. It contains tools for excitation signal design, data preprocessing, parameter estimation, model and error presentation, model validation, simulation, archivation and documentation. This paper illustrates the use of many functions of the toolbox on the example of the identification of a model of a flexible robot arm.