Multiscale model reduction methods for deterministic and stochastic partial differential equations

Partial differential equations (PDEs) with multiscale coefficients are very difficult to solve due to the wide range of scales in the solutions. In the thesis, we propose some efficient numerical methods for both deterministic and stochastic PDEs based on the model reduction technique.

For the deterministic PDEs, the main purpose of our method is to derive an effective equation for the multiscale problem. An essential ingredient is to decompose the harmonic coordinate into a smooth part and a highly oscillatory part of which the magnitude is small. Such a decomposition plays a key role in our construction of the effective equation. We show that the solution to the effective equation is smooth, and could be resolved on a regular coarse mesh grid. Furthermore, we provide error analysis and show that the solution to the effective equation plus a correction term is close to the original multiscale solution.

For the stochastic PDEs, we propose the model reduction based data-driven stochastic method and multilevel Monte Carlo method. In the multiquery, setting and on the assumption that the ratio of the smallest scale and largest scale is not too small, we propose the multiscale data-driven stochastic method. We construct a data-driven stochastic basis and solve the coupled deterministic PDEs to obtain the solutions. For the tougher problems, we propose the multiscale multilevel Monte Carlo method. We apply the multilevel scheme to the effective equations and assemble the stiffness matrices efficiently on each coarse mesh grid. In both methods, the $\KL$ expansion plays an important role in extracting the main parts of some stochastic quantities.

For both the deterministic and stochastic PDEs, numerical results are presented to demonstrate the accuracy and robustness of the methods. We also show the computational time cost reduction in the numerical examples.

Uniformly convergent finite element methods for singularly perturbed parabolic partial differential equations

This thesis is concerned with uniformly convergent finite element methods for numerically solving singularly perturbed parabolic partial differential equations in one space variable. First, we use Petrov-Galerkin finite element methods to generate three schemes for such problems, each of these schemes uses exponentially fitted elements in space. Two of them are lumped and the other is non-lumped. On meshes which are either arbitrary or slightly restricted, we derive global energy norm and L2 norm error bounds, uniformly in the diffusion parameter. Under some reasonable global assumptions together with realistic local assumptions on the solution and its derivatives, we prove that these exponentially fitted schemes are locally uniformly convergent, with order one, in a discrete L∞norm both outside and inside the boundary layer. We next analyse a streamline diffusion scheme on a Shishkin mesh for a model singularly perturbed parabolic partial differential equation. The method with piecewise linear space-time elements is shown, under reasonable assumptions on the solution, to be convergent, independently of the diffusion parameter, with a pointwise accuracy of almost order 5/4 outside layers and almost order 3/4 inside the boundary layer. Numerical results for the above schemes are presented. Finally, we examine a cell vertex finite volume method which is applied to a model time-dependent convection-diffusion problem. Local errors away from all layers are obtained in the l2 seminorm by using techniques from finite element analysis.

Automated synthesis procedure of RF discrete tuning differential capacitance circuits

The paper presents a RFDSCA automated synthesis procedure. This algorithm determines several RFDSCA circuits from the top-level system specifications all with the same maximum performance. The genetic synthesis tool optimizes a fitness function proportional to the RFDSCA quality factor and uses the epsiv-concept and maximin sorting scheme to achieve a set of solutions well distributed along a non-dominated front. To confirm the results of the algorithm, three RFDSCAs were simulated in SpectreRF and one of them was implemented and tested. The design used a 0.25 mum BiCMOS process. All the results (synthesized, simulated and measured) are very close, which indicate that the genetic synthesis method is a very useful tool to design optimum performance RFDSCAs.

Invariant discretizations of partial differential equations

Un algorithme permettant de discrétiser les équations aux dérivées partielles (EDP) tout en préservant leurs symétries de Lie est élaboré. Ceci est rendu possible grâce à l'utilisation de dérivées partielles discrètes se transformant comme les dérivées partielles continues sous l'action de groupes de Lie locaux. Dans les applications, beaucoup d'EDP sont invariantes sous l'action de transformations ponctuelles de Lie de dimension infinie qui font partie de ce que l'on désigne comme des pseudo-groupes de Lie. Afin d'étendre la méthode de discrétisation préservant les symétries à ces équations, une discrétisation des pseudo-groupes est proposée. Cette discrétisation a pour effet de transformer les symétries ponctuelles en symétries généralisées dans l'espace discret. Des schémas invariants sont ensuite créés pour un certain nombre d'EDP. Dans tous les cas, des tests numériques montrent que les schémas invariants approximent mieux leur équivalent continu que les différences finies standard.

Upwind solution of singular differential equations arising from steady channel flows

We study ordinary nonlinear singular differential equations which arise from steady conservation laws with source terms. An example of steady conservation laws which leads to those scalar equations is the Saint–Venant equations. The numerical solution of these scalar equations is sought by using the ideas of upwinding and discretisation of source terms. Both the Engquist–Osher scheme and the Roe scheme are used with different strategies for discretising the source terms.

Scale-invariant moving finite elements for nonlinear partial differential equations in two dimensions

A scale-invariant moving finite element method is proposed for the adaptive solution of nonlinear partial differential equations. The mesh movement is based on a finite element discretisation of a scale-invariant conservation principle incorporating a monitor function, while the time discretisation of the resulting system of ordinary differential equations is carried out using a scale-invariant time-stepping which yields uniform local accuracy in time. The accuracy and reliability of the algorithm are successfully tested against exact self-similar solutions where available, and otherwise against a state-of-the-art h-refinement scheme for solutions of a two-dimensional porous medium equation problem with a moving boundary. The monitor functions used are the dependent variable and a monitor related to the surface area of the solution manifold. (c) 2005 IMACS. Published by Elsevier B.V. All rights reserved.

Advances in sequential data assimilation and numerical weather forecasting: an Ensemble Transform Kalman-Bucy Filter, a study on clustering in deterministic Ensemble Square Root Filters, and a test of a new time stepping scheme in an atmospheric model

This dissertation deals with aspects of sequential data assimilation (in particular ensemble Kalman filtering) and numerical weather forecasting. In the first part, the recently formulated Ensemble Kalman-Bucy (EnKBF) filter is revisited. It is shown that the previously used numerical integration scheme fails when the magnitude of the background error covariance grows beyond that of the observational error covariance in the forecast window. Therefore, we present a suitable integration scheme that handles the stiffening of the differential equations involved and doesn’t represent further computational expense. Moreover, a transform-based alternative to the EnKBF is developed: under this scheme, the operations are performed in the ensemble space instead of in the state space. Advantages of this formulation are explained. For the first time, the EnKBF is implemented in an atmospheric model. The second part of this work deals with ensemble clustering, a phenomenon that arises when performing data assimilation using of deterministic ensemble square root filters in highly nonlinear forecast models. Namely, an M-member ensemble detaches into an outlier and a cluster of M-1 members. Previous works may suggest that this issue represents a failure of EnSRFs; this work dispels that notion. It is shown that ensemble clustering can be reverted also due to nonlinear processes, in particular the alternation between nonlinear expansion and compression of the ensemble for different regions of the attractor. Some EnSRFs that use random rotations have been developed to overcome this issue; these formulations are analyzed and their advantages and disadvantages with respect to common EnSRFs are discussed. The third and last part contains the implementation of the Robert-Asselin-Williams (RAW) filter in an atmospheric model. The RAW filter is an improvement to the widely popular Robert-Asselin filter that successfully suppresses spurious computational waves while avoiding any distortion in the mean value of the function. Using statistical significance tests both at the local and field level, it is shown that the climatology of the SPEEDY model is not modified by the changed time stepping scheme; hence, no retuning of the parameterizations is required. It is found the accuracy of the medium-term forecasts is increased by using the RAW filter.

A geometric mass-preserving redistancing scheme for the level set function

In this paper we describe and evaluate a geometric mass-preserving redistancing procedure for the level set function on general structured grids. The proposed algorithm is adapted from a recent finite element-based method and preserves the mass by means of a localized mass correction. A salient feature of the scheme is the absence of adjustable parameters. The algorithm is tested in two and three spatial dimensions and compared with the widely used partial differential equation (PDE)-based redistancing method using structured Cartesian grids. Through the use of quantitative error measures of interest in level set methods, we show that the overall performance of the proposed geometric procedure is better than PDE-based reinitialization schemes, since it is more robust with comparable accuracy. We also show that the algorithm is well-suited for the highly stretched curvilinear grids used in CFD simulations. Copyright (C) 2010 John Wiley & Sons, Ltd.

A Direct Numerov Sixth-order Numerical Scheme to Accurately Solve the Unidimensional Poisson Equation with Dirichlet Boundary Conditions

In this article, we present an analytical direct method, based on a Numerov three-point scheme, which is sixth order accurate and has a linear execution time on the grid dimension, to solve the discrete one-dimensional Poisson equation with Dirichlet boundary conditions. Our results should improve numerical codes used mainly in self-consistent calculations in solid state physics.

A DISCRETIZATION SCHEME FOR AN ONE-DIMENSIONAL REACTION-DIFFUSION EQUATION WITH DELAY AND ITS DYNAMICS

The goal of this paper is to present an approximation scheme for a reaction-diffusion equation with finite delay, which has been used as a model to study the evolution of a population with density distribution u, in such a way that the resulting finite dimensional ordinary differential system contains the same asymptotic dynamics as the reaction-diffusion equation.

Differential encoding technique for multi-antenna systems with correlated Rayleigh fading channels

Differential space-time modulation (DSTM) techniques developed for multi-antenna systems allow the receiver to detect the transmitted signal without the knowledge of the fading channels. It can be viewed as an extension of differential phase-shift keying (DPSK) in single antenna systems. In this paper, we derived the pairwise error probability upper bound of differential space-time coded systems with spatially correlated Rayleigh fading channels. Based on the performance analysis, we develop a novel DSTM scheme which can exploit the spatial correlation in the fading channels. It is found that by carefully designing the initial transmitted signal matrix, the performance of the differential space-time coded systems can be significantly improved.

A generalized data windowing scheme for adaptive conjugate gradient algorithms

The performance of the modified adaptive conjugate gradient (CG) algorithms based on the iterative CG method for adaptive filtering is highly related to the ways of estimating the correlation matrix and the cross-correlation vector. The existing approaches of implementing the CG algorithms using the data windows of exponential form or sliding form result in either loss of convergence or increase in misadjustment. This paper presents and analyzes a new approach to the implementation of the CG algorithms for adaptive filtering by using a generalized data windowing scheme. For the new modified CG algorithms, we show that the convergence speed is accelerated, the misadjustment and tracking capability comparable to those of the recursive least squares (RLS) algorithm are achieved. Computer simulations demonstrated in the framework of linear system modeling problem show the improvements of the new modifications.

Channel estimation scheme for 3.9G wireless communication systems using RLS algorithm

Main challenges for a terminal implementation are efficient realization of the receiver, especially for channel estimation (CE) and equalization. In this paper, training based recursive least square (RLS) channel estimator technique is presented for a long term evolution (LTE) single carrier-frequency division multiple access (SC-FDMA) wireless communication system. This CE scheme uses adaptive RLS estimator which is able to update parameters of the estimator continuously, so that knowledge of channel and noise statistics are not required. Simulation results show that the RLS CE scheme with 500 Hz Doppler frequency has 3 dB better performances compared with 1.5 kHz Doppler frequency.

A new sliding mode-based learning control scheme

 A new sliding mode-based learning control scheme for a class of SISO dynamic systems is developed in this paper. It is seen that, based on the most recent information on the closed-loop stability, a recursive learning chattering-free sliding mode controller can be designed to drive the closed-loop dynamics to reach the sliding mode surface in a finite time, on which the desired closed-loop dynamics with the zero-error convergence can be achieved.

Rational expectations and the service of public debt: a recursive approach

This paper uses dynamic programming to study the time consistency of optimal macroeconomic policy in economies with recurring public deficits. To this end, a general equilibrium recursive model introduced in Chang (1998) is extended to include govemment bonds and production. The original mode! presents a Sidrauski economy with money and transfers only, implying that the need for govemment fmancing through the inflation tax is minimal. The extended model introduces govemment expenditures and a deficit-financing scheme, analyzing the SargentWallace (1981) problem: recurring deficits may lead the govemment to default on part of its public debt through inflation. The methodology allows for the computation of the set of alI sustainable stabilization plans even when the govemment cannot pre-commit to an optimal inflation path. This is done through value function iterations, which can be done on a computeI. The parameters of the extended model are calibrated with Brazilian data, using as case study three Brazilian stabilization attempts: the Cruzado (1986), Collor (1990) and the Real (1994) plans. The calibration of the parameters of the extended model is straightforward, but its numerical solution proves unfeasible due to a dimensionality problem in the algorithm arising from limitations of available computer technology. However, a numerical solution using the original algorithm and some calibrated parameters is obtained. Results indicate that in the absence of govemment bonds or production only the Real Plan is sustainable in the long run. The numerical solution of the extended algorithm is left for future research.