Naslov
Analysis of Coupled Dynamical Systems

Autori
Dogančić, Bruno ; Jokić, Marko

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
Third Annual PhD Workshop Book of Abstracts / Parunov, Joško ; Zovko Brodarac, Zdenka ; Bauer, Branko ; Duić, Neven ; Jokić, Andrej ; Landek, Darko ; Lisjak, Dragutin ; Lulić, Zoran ; Majetić, Dubravko ; Matijević, Božidar ; Runje, Biserka ; Sorić, Jurica ; Terze, Zdravko - Zagreb : Fakultet strojarstva i brodogradnje Sveučilišta u Zagrebu ; Metalurški fakultet Sveučilišta u Zagrebu, 2017, 17-17

ISBN
978-953-7738-53-2

Skup
Third Annual PhD Workshop

Mjesto i datum
Zagreb, Hrvatska, 30.07.2017

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Domaća recenzija

Ključne riječi
coupled dynamical systems, a priori error estimation, uncertainty modeling

Sažetak
Introduction Dynamical systems are becoming more complex and thereby demands in terms of robustness and effi ciency are set high when synthesizing such dynamical systems. Common paradigm, when coupled dynamical systems are modelled, includes mathematical modeling of each subsystem described with system of constitutive relations. These relations usually yield system of partial diff erential equations (PDE). PDE are later approximated with discretization tools such as fi nite element method (FEM) to achieve system of ordinary diff erential equations (ODE). Discretized subsystems are then coupled into complex dynamical systems. When observing such systems one has to answer questions regarding validity and accuracy of mathematical model. Since each discretized subsystem carries approximation error the question arises: ”When is discretization of each subsystem good enough to maintain accuracy and validity of resulting coupled dynamical system?“. Aims Robust control theory demands uncertainty modelling. Aim is to formulate uncertainty model of an individual subsystem that describes discretization error. When doing so, dynamical systems will be analyzed using interconnected plants consisting of linear time invariant plants (LTI) with feedback. Feedback of the system includes uncertainties. The fact that each subsystem is coupled with other subsystems will (hopefully) be used to reduce conservatism and provide bett er uncertainty model. Methods In modern robust control theory, integral quadratic constraints (IQC) are often used for modelling wide variety of uncertainty classes. Together with other mathematical tools and models, such as linear matrix inequalities (LMI), dissipation theory and convex optimization, IQC will be used to model uncertainty due to discretization. Expected scientifi c contribution Resulting a- priori error estimator should provide effi cient and robust tool for analyzing discretized coupled dynamical systems. This method is effi cient since computationally undemanding coarse mesh can be used for discretization and linear systems of equation are solved. The resulting model (LTI plus uncertainties) is suitable for robust control since the uncertainty model is less conservative.

Izvorni jezik
Engleski

POVEZANOST RADA