Les simulations à grande dimension jouent un rôle crucial dans l'étude d'une grande variété de phénomènes physiques complexes, entraînant souvent des demandes écrasantes sur les ressources informatiques. La gestion de ces demandes constitue la principale motivation pour la réduction du modèle : produire des modèles de commande réduite plus simples, qui permettent une simulation plus rapide et moins coûteuse tout en se rapprochant avec précision du comportement du modèle d'origine. La présence des systèmes avec multiples entrées et multiples sorties (MIMO) rend le processus de réduction encore plus difficile. Dans cette thèse, nous nous intéressons aux méthodes de réduction de modèles à grande dimension en utilisant la projection sur des sous-espaces de Krylov tangentielles. Nous nous penchons sur le développement de techniques qui utilisent l'interpolation tangentielle. Celles-ci présentent une alternative efficace et intéressante à la troncature équilibrée qui est considérée comme référence dans le domaine et tout particulièrement la réduction pour les systèmes linéaire à temps invariants. Enfin, une attention particulière sera portée sur l'élaboration de nouveaux algorithmes efficaces et sur l'application à des problèmes pratiques. / Large-scale simulations play a crucial role in the study of a great variety of complex physical phenomena, leading often to overwhelming demands on computational resources. Managing these demands constitutes the main motivation for model reduction : produce simpler reduced-order models, which allow for faster and cheaper simulation while accurately approximating the behaviour of the original model. The presence of multiple inputs and outputs (MIMO) systems, makes the reduction process even more challenging. In this thesis we are interested in methods of reducing large-scale models, using projection on tangential Krylov subspaces. We are looking at the development of techniques using tangential interpolation. These present an effective and interesting alternative to the balanced truncation which is considered as a reference in the field and especially for the reduction of linear time invariant systems. Finally, special attention will be focused on the development of new efficient algorithms and application to practical problems.

A multiple-point Arnoldi method is derived for model reduction of computational fluid dynamic systems. By choosing the number of frequency interpolation points and the number of Arnoldi vectors at each frequency point, the user can select the accuracy and range of validity of the resulting reduced-order model while balancing computational expense. The multiple-point Arnoldi approach is combined with a singular value decomposition approach similar to that used in the proper orthogonal decomposition method. This additional processing of the basis allows a further reduction in the number of states to be obtained, while retaining a significant computational cost advantage over the proper orthogonal decomposition. Results are presented for a supersonic diffuser subject to mass flow bleed at the wall and perturbations in the incoming flow. The resulting reduced-order models capture the required dynamics accurately while providing a significant reduction in the number of states. The reduced-order models are used to generate transfer function data, which are then used to design a simple feedforward controller. The controller is shown to work effectively at maintaining the average diffuser throat Mach number. / Singapore-MIT Alliance (SMA)

In this paper we present an approach to the nonlinear model reduction based on representing the nonlinear system with a piecewise-linear system and then reducing each of the pieces with a Krylov projection. However, rather than approximating the individual components to make a system with exponentially many different linear regions, we instead generate a small set of linearizations about the state trajectory which is the response to a 'training input'. Computational results and performance data are presented for a nonlinear circuit and a micromachined fixed-fixed beam example. These examples demonstrate that the macromodels obtained with the proposed reduction algorithm are significantly more accurate than models obtained with linear or the recently developed quadratic reduction techniques. Finally, it is shown tat the proposed technique is computationally inexpensive, and that the models can be constructed 'on-the-fly', to accelerate simulation of the system response. / Singapore-MIT Alliance (SMA)

The thesis treats model reduction for linear time-varying systems. Time-varying models appear in many fields, including power systems, chemical engineering, aeronautics, and computational science. They can also be used for approximation of time-invariant nonlinear models. Model reduction is a topic that deals with simplification of complex models. This is important since it facilitates analysis and synthesis of controllers. The thesis consists of two parts. The first part provides an introduction to the topics of time-varying systems and model reduction. Here, notation, standard results, examples, and some results from the second part of the thesis are presented. The second part of the thesis consists of four papers. In the first paper, we study the balanced truncation method for linear time-varying state-space models. We derive error bounds for the simplified models. These bounds are generalizations of well-known time-invariant results, derived with other methods. In the second paper, we apply balanced truncation to a high-order model of a diesel exhaust catalyst. Furthermore, we discuss practical issues of balanced truncation and approximative discretization. In the third paper, we look at frequency-domain analysis of linear time-periodic impulse-response models. By decomposing the models into Taylor and Fourier series, we can analyze convergence properties of different truncated representations. In the fourth paper, we use the frequency-domain representation developed in the third paper, the harmonic transfer function, to generalize Bode's sensitivity integral. This result quantifies limitations for feedback control of linear time-periodic systems. / QC 20120206

In this thesis the stability of generic boundary layer flows is studied from a global viewpoint using optimization methods. Global eigenmodes of the incompressible linearized Navier-Stokes equations are computed using the Krylov subspace Arnoldi method. These modes serve as a tool both to study asymptotic stability and as a reduced basis to study transient growth. Transient growth is also studied using adjoint iterations. The knowledge obtained from the stability analysis is used to device systematic feedback control in the Linear Quadratic Gaussian framework. The dynamics is assumed to be described by the linearized Navier-Stokes equations. Actuators and sensors are designed and a Kalman filtering technique is used to reconstruct the unknown flow state from noisy measurements. This reconstructed flow state is used to determine the control feedback which is applied to the Navier-Stokes equations through properly designed actuators. Since the control and estimation gains are obtained through an optimization process, and the Navier-Stokes equations typically forms a very high-dimensional system when discretized there is an interest in reducing the complexity of the equations. A standard method to construct a reduced order model is to perform a Galerkin projection of the full equations onto the subspace spanned by a suitable set of vectors, such as global eigenmodes and balanced truncation modes. / QC 20100924

Flow control of transitional shear flows is investigated by means of numerical simulations. The attenuation of three-dimensional wavepackets of Tollmien-Schlichting (TS) and streaks in the boundary layer is obtained using active control in combination with localised sensors and actuators distributed near the rigid wall. Due to the dimensions of the discretized Navier-Stokes operator, reduced-order models are identified, preserving the dynamics between the inputs and the outputs of the system. Balanced realizations of the system are computed using balanced truncation and system identification. We demonstrate that the energy growth of the perturbations is substantially and efficiently mitigated, using relatively few sensors and actuators. The robustness of the controller is analysed by varying the number of actuators and sensors, the Reynolds number, the pressure gradient and by investigating the nonlinear, transitional case. We show that delay of the transition from laminar to turbulent flow can be achieved despite the fully linear approach. This configuration can be reproduced in experiments, due to the localisation of sensing and actuation devices. The closed-loop system has been investigated for the corresponding twodimensional case by using full-dimensional optimal controllers computed by solving an iterative optimisation based on the Lagrangian approach. This strategy allows to compare the results achieved using open-loop model reduction with model-free controllers. Finally, a parametric analysis of the actuators/ sensors placement is carried-out to deepen the understanding of the inherent dynamics of the closed-loop. The distinction among two different classes of controllers – feedforward and feedback controllers - is highlighted. A second shear flow, a confined turbulent jet, is investigated using particle image velocimetry (PIV) measurements. Proper orthogonal decomposition (POD) modes and Koopman modes via dynamic mode decomposition (DMD) are computed and analysed for understanding the main features of the flow. The frequencies related to the dominating mechanisms are identified; the most energetic structures show temporal periodicity. / <p>QC 20130207</p>

Studies of spatial memory provide valuable insight into more general mnemonic functions, for by observing the activity of cells such as place cells, one can follow a subject’s dynamic representation of a changing environment. I investigate how place cells resolve conflicting neuronal input signals by developing computational models that integrate synaptic inputs on two scales. First, I construct reduced models of morphologically accurate neurons that preserve neuronal structure and the spatial specificity of inputs. Second, I use a parallel implementation to examine the dynamics among a network of interconnected place cells. Both models elucidate possible roles for the inputs and mechanisms involved in spatial memory.

This thesis proposes a model reduction technique for nonlinear dynamical systems based upon combining Proper Orthogonal Decomposition (POD) and a new method, called the Discrete Empirical Interpolation Method (DEIM). The popular method of Galerkin projection with POD basis reduces dimension in the sense that far fewer variables are present, but the complexity of evaluating the nonlinear term generally remains that of the original problem. DEIM, a discrete variant of the approach from [11], is introduced and shown to effectively overcome this complexity issue. State space error estimates for POD-DEIM reduced systems are also derived. These [Special characters omitted.] error estimates reflect the POD approximation property through the decay of certain singular values and explain how the DEIM approximation error involving the nonlinear term comes into play. An application to the simulation of nonlinear miscible flow in a 2-D porous medium shows that the dynamics of a complex full-order system of dimension 15000 can be captured accurately by the POD-DEIM reduced system of dimension 40 with a factor of [Special characters omitted.] (1000) reduction in computational time.

<p>A new method of stabilizing low-order, proper orthogonal decomposition based reduced-order models of the Navier Stokes equations is proposed. Unlike traditional approaches, this method does not rely on empirical turbulence modeling or modification of the Navier-Stokes equations. It provides spatial basis functions different from the usual proper orthogonal decomposition basis function in that, in addition to optimally representing the solution, the new proposed basis functions also provide stable reduced-order models. The proposed approach is illustrated with two test cases: two-dimensional flow inside a square lid-driven cavity and a two-dimensional mixing layer.</p> / Dissertation

This dissertation discusses the predictability of the atmosphere-ocean climate system on interannual and decadal timescales. We investigate the extent to which the atmospheric internal variability (weather noise) can cause climate prediction to lose skill; and we also look for the oceanic processes that contribute to the climate predictability via interaction with the atmosphere. First, we develop a framework for assessing the predictability of a linear stochastic system. Based on the information of deterministic dynamics and noise forcing, various predictability measures are defined and new predictability-analysis tools are introduced. For the sake of computational efficiency, we also discuss the formulation of a low-order model within the context of four reduction methods: modal, EOF, most predictable pattern, and balanced truncation. Subsequently, predictabilities of two specific physical systems are investigated within such framework. The first is a mixed layer model of SST with focus on the effect of oceanic advection.Analytical solution of a one-dimensional model shows that even though advection can give rise to a pair of low-frequency normal modes, no enhancement in the predictability is found in terms of domain averaged error variance. However, a Predictable Component Analysis (PrCA) shows that advection can play a role in redistributing the predictable variance. This analytical result is further tested in a more realistic two-dimensional North Atlantic model with observed mean currents. The second is a linear coupled model of tropical Atlantic atmosphere-ocean system. Eigen-analysis reveals that the system has two types of coupled modes: a decadal meridional mode and an interannual equatorial mode. The meridional mode, which manifests itself as a dipole pattern in SST, is controlled by thermodynamic feedback between wind, latent heat flux, and SST, and modified by ocean heat transport. The equatorial mode, which manifests itself as an SST anomaly in the eastern equatorial basin, is dominated by dynamic feedback between wind, thermocline, upwelling, and SST. The relative strength of thermodynamic vs dynamic feedbacks determines the behavior of the coupled system, and enables the tropical Atlantic variability to be more predictable than the passive-ocean scenario.