Dynamical modeling with application to friction phenomena
A. Hornstein · 2005
The thesis deals with dynamical modeling as an alternative approach to the conventional way of modeling with static models. From the perspective of generalized synchronization the main concepts of dynamical modeling are developed. As an important subclass of dynamical models, Recurrent Neural Networks are investigated and several algorithms for improved modeling are presented. In the last part of the thesis dynamical modeling is applied to friction phenomena. Data from friction experiments is used to model the diverse apects of friction in the sliding and pre-sliding regime.