<p>Diffusion is a phenomenon in which particles move from regions of higher density to regions of lower density. Many physical systems, in fields as diverse as plant biology and finance, are known to involve diffusion phenomena. Typically, diffusion systems are modeled by partial differential equations (PDEs), which include certain parameters. These parameters characterize a given diffusion system. Therefore, for both modeling and simulation of a diffusion system, one has to either know or determine these parameters. Moreover, as PDEs are infinite order dynamic systems, for computational purposes one has to approximate them by a finite order model. In this thesis, we investigate these two issues of model reduction and parameter estimation by considering certain specific cases of heat diffusion systems. </p><p>We first address model reduction by considering two specific cases of heat diffusion systems. The first case is a one-dimensional heat diffusion across a homogeneous wall, and the second case is a two-dimensional heat diffusion across a homogeneous rectangular plate. In the one-dimensional case we construct finite order approximations by using some well known PDE solvers and evaluate their effectiveness in approximating the true system. We also construct certain other alternative approximations for the one-dimensional diffusion system by exploiting the different modal structures inherently present in it. For the two-dimensional heat diffusion system, we construct finite order approximations first using the standard finite difference approximation (FD) scheme, and then refine the FD approximation by using its asymptotic limit.</p><p>As for parameter estimation, we consider the same one-dimensional heat diffusion system, as in model reduction. We estimate the parameters involved, first using the standard batch estimation technique. The convergence of the estimates are investigated both numerically and theoretically. We also estimate the parameters of the one-dimensional heat diffusion system recursively, initially by adopting the standard recursive prediction error method (RPEM), and later by using two different recursive algorithms devised in the frequency domain. The convergence of the frequency domain recursive estimates is also investigated. </p>

In diese Dissertation werden lineare retardierte Differentialgleichungen (DDEs) und deren Lösungsoperatoren untersucht. Wir stellen eine neue Methode vor, mit der die Lösungsoperatoren für autonome und nicht-autonome DDEs bestimmt werden. Die neue Methode basiert auf dem Pfadintegralformalismus, der aus der Quantenmechanik und von der Analyse stochastischer Differentialgleichungen bekannt ist. Es zeigt sich, dass die Lösung eines Delaysystems zum Zeitpunkt t durch die Integration aller möglicher Pfade von der Anfangsbedingung bis zur Zeit t gebildet werden kann. Die Pfade bestehen dabei aus verschiedenen Schritten unterschiedlicher Längen und Gewichte. Für skalare autonome DDEs können analytische Ausdrücke des Lösungsoperators in der Literatur gefunden werden, allerdings existieren keine für nicht-autonome oder höherdimensionale DDEs. Mithilfe der neuen Methode werden wir die Lösungsoperatoren der genannten DDEs aufstellen und zusätzlich auf Delaysysteme mit mehreren Delaytermen erweitern. Dabei bestätigen wir unsere Ergebnisse sowohl analytisch wie auch numerisch. Die gewonnenen Lösungsoperatoren verwenden wir anschließend zur Stabilitätsanalyse periodischer Delaysysteme. Es werden zwei neue Verfahren präsentiert, die mithilfe des Lösungsoperators den transformierten Monodromieoperator des Delaysystems nähern und daraus die Stabilität bestimmen können. Beide neue Verfahren sind spektrale Methoden für autonome sowie nicht-autonome Delaysysteme und haben keine Einschränkungen wie bei der bekannten Chebyshev-Kollokationsmethode oder der Chebyshev-Polynomentwicklung. Die beiden bisherigen Verfahren beschränken sich auf Delaysysteme mit rationalem Verhältnis zwischen Periode und Delay. Außerdem werden wir eine bereits bekannte Methode erweitern und zu einer spektralen Methode für periodische nicht-autonome Delaysysteme entwickeln. Wir bestätigen alle drei neue Verfahren numerisch. Damit werden in dieser Dissertation drei neue spektrale Verfahren zur Stabilitätsanalyse periodischer Delaysysteme vorgestellt. / In this thesis linear delay differential equations (DDEs) and its solutions operators are studied. We present a new method to calculate the solution operators for autonomous and non-autonomous DDEs. The new method is related to the path integral formalism, which is known from quantum mechanics and the analysis of stochastic differential equations. It will be shown that the solution of a time delay system at time t can be constructed by integrating over all paths from the initial condition to time t. The paths consist of several steps with different lengths and weights. Analytic expressions for the solution operator for scalar autonomous DDEs can be found in the literature but no results exist for non-autonomous or high dimensional DDEs. With the help of the new method we can calculate the solution operators for such DDEs and for time delay systems with several delay terms. We verify our results analytically and numerically. We use the obtained solution operators for the stability analysis of periodic time delay systems. Two new methods will be presented to approximate the transformed monodromy operator with the help of the solution operator and to get the stability. Both new methods are spectral methods for autonomous and non-autonomous delay systems and have no limitations like the known Chebyshev collocation method or Chebyshev polynomial expansion. Both previously known methods are limited to time delay systems with a rational relation between period and delay. Furthermore we will extend a known method to a spectral method for non-autonomous time delay systems. We verify all three new methods numerically. Hence, in this thesis three new spectral methods for the stability analysis of periodic time delay systems are presented.