System Identification and Parameter Estimation

System identification deals with the estimation of physical properties of (dynamic) systems from measured data (inputs and outputs), which is essential for the understanding of system behavior. System identification is particularly useful for biological systems where physical entities are generally not known a priori. The course provides mathematical a background of system identification in both time domain and frequency domain with a strong focus on practical application. Major topics to be studied are: spectral densities, nonparametric black-box models, discrete time domain models (parametric grey-box), closed loop systems (utilizing feedback), translation of transfer functions into physical parameters, optimization routines and design of perturbation signals. During the course examples from both technical and physiological systems (i.e. estimate the behavior of a driver or the dynamics of a human joint) will be discussed.