|
|
|
Abstract:
|
One of the core issues in the area of Systems and Control is the creation of reliable dynamical models of physical and chemical processes, often to be used for the purpose of model-based control design. Due to increasing demands with respect to performance and efficiency of the resulting control systems, there is a constant need to improve this modeling process. This thesis presents a sound modeling framework, that is based on the model class of Linear Parameter Varying (LPV) systems, motivated by the wide representation capabilities of these models and a well worked-out and industrially reputed LPV control synthesis theory. Using this framework a number of promising approaches is developed for the estimation of LPV models, based on experimental data. A key concept in this system identification methodology is the use of series-expansion descriptions of LPV systems in terms of orthonormal basis functions.
|
