No description available.

No description available.

I Sverige så har användningen av inhyrd arbetskraft under de senaste åren starkt ökat (Handelns utredningsinstitut, 2010b). Användningen av inhyrd arbetskraft kan antas vara en vilja från företaget att anpassa sig efter teorin om den flexibla organisationen (Olofsdotter, 2008:7). En funktionellt flexibel kärna av arbetskraft skall kompletteras med en periferi av numeriskt flexibel arbetskraft (Atkinson & Gregory, 1986:13). Det saknas dock studier på hur företagets inre maktförhållanden påverkas av en sådan organisering. Syftet med denna studie är att beskriva och kartlägga hur en organisations maktstrukturer påverkas av uppdelningen av organisationens arbetskraft enligt kategorierna funktionellt- och numerisktflexibel arbetskraft. För att uppnå detta syfte så har en kvalitativ studie, bestående av observationer och intervjuer, genomförts på ett företag inom den marina tillverkningsindustrin. Studiens resultat visar på att företagets informella maktstrukturer ändras till följd av användningen av inhyrd arbetskraft. Produktionen i det studerade företaget är organiserat så att verkstadsarbetare ifrån de båda arbetskraftskategorierna, funktionellt och numeriskt flexibla, utvecklar ett ömsesidigt beroende. Det ömsesidiga beroendet gör att den statusökning som den ordinarie personalen skulle få, som en följd av att utgöra den funktionellt flexibla delen av företaget, uteblir. Den uteblivna statusförhöjningen för de anställda leder i sin tur till ett känslomässigt avståndstagande till företaget. Studien har därmed visat på att uppdelningen av arbetskraften i två kategorier får inverkan på organisationens sociala strukturer. En inverkan som tidigare lämnats outforskad.

In this masters’ thesis report we present a graph theoretical method used for automatic bilingual lexicon acquisition with parallel texts. We analyze the concept of graph similarity and give an interpretation, of the parallel texts, connected to the vector space model. We represent the parallel texts by a directed, tripartite graph and from here use the corresponding adjacency matrix, A, to compute the similarity of the graph. By solving the eigenvalue problem ρS = ASAT + ATSA we obtain the self-similarity matrix S and the Perron root ρ. A rank k approximation of the self-similarity matrix is computed by implementations of the singular value decomposition and the non-negative matrix factorization algorithm GD-CLS. We construct an algorithm in order to extract the bilingual lexicon from the self-similarity matrix and apply a statistical model to estimate the precision, the correctness, of the translations in the bilingual lexicon. The best result is achieved with an application of the vector space model with a precision of about 80 %. This is a good result and can be compared with the precision of about 60 % found in the literature.

<p>Consider a mixture of two immiscible, incompressible fluids e.g. oil and water. Since the fluids do not mix, an interface between the two fluids will form and move in time. The motion of the two fluids can be modelled by the incompressible Navier-Stokes equations for two phase flow with surface tension together with a representation of the moving interface. The parameters in the Navier-Stokes equations will depend on the position and other properties of the interface. The interface should move with the velocity of the flow at the interface. Since the fluids are incompressible, the density of each fluid is constant. Mass conservation then implies that the volume occupied by each of the two fluids should not change with time. The object of this thesis has been to develop a new numerical method to simulate incompressible two phase flow accurately that conserves mass and volume of each fluid correctly.</p><p>Numerical simulations of incompressible two phase flow with surface tension have been a challenge for many years. Several methods have been developed and used prior to the work presented in this thesis. The two most commonly used methods are volume of fluid methods and level set methods. There are advantages and disadvantages of both of the methods.</p><p>In volume of fluid methods the interface is represented by a discontinuity of a globally defined function. Because of the discontinuity it is hard both to move the interface as well as to calculate properties of the interface such as curvature. Specially designed methods have to be used, and all these methods are low order accurate. Volume of fluid methods do however conserve the volumes of the two fluids correctly.</p><p>In level set methods the interface is represented by the zero contour of the globally defined signed distance function. This function is smooth across the interface. Since the function is smooth, standard methods for partial differential equations can be used to advect the interface accurately. A reinitialization is however needed to make sure that the level set function remains a signed distance function. During this process the zero contour might move slightly. Because of this, the volume conservation of the method becomes poor.</p><p>In this thesis we present a new level set method. The method is designed such that the volume of each fluid is conserved, at least approximately. The interface is represented by the 0.5 contour of a regularized characteristic function. As for standard level set methods, the interface is moved first by an advective step, and then reinitialized. Unlike traditional level set methods, we can formulate the reinitialization as a conservation law. Conservative methods can then be used to move and to reinitialize the level set function numerically. Since the level set function is a regularized characteristic function, we can expect good conservation of the volume bounded by the interface.</p><p>The method is discretized using both finite differences and finite elements. Uniform and adaptive grids are used in both two and three space dimensions. Good convergence as well as volume conservation is observed. Theoretical studies are performed to investigate the conservation and the computational time needed for reinitialization.</p>

<p>This thesis concerns the approximation of optimally controlled partial differential equations for applications in optimal design and reconstruction. Such optimal control problems are often ill-posed and need to be regularized to obtain good approximations. We here use the theory of the corresponding Hamilton-Jacobi-Bellman equations to construct regularizations and derive error estimates for optimal design problems. The constructed Pontryagin method is a simple and general method where the first, analytical, step is to regularize the Hamiltonian. Next its stationary Hamiltonian system, a nonlinear partial differential equation, is computed efficiently with the Newton method using a sparse Jacobian. An error estimate for the difference between exact and approximate objective functions is derived, depending only on the difference of the Hamiltonian and its finite dimensional regularization along the solution path and its<em> L</em>² projection, i.e. not on the difference of the exact and approximate solutions to the Hamiltonian systems. In the thesis we present solutions to applications such as optimal design and reconstruction of conducting materials and elastic structures.</p>

<p>This thesis concerns the numerical simulation of creeping rays and their contribution to high frequency scattering problems.</p><p>Creeping rays are a type of diffracted rays which are generated at the shadow line of the scatterer and propagate along geodesic paths on the scatterer surface. On a perfectly conducting convex body, they attenuate along their propagation path by tangentially shedding diffracted rays and losing energy. On a concave scatterer, they propagate on the surface and importantly, in the absence of dissipation, experience no attenuation. The study of creeping rays is important in many high frequency problems, such as design of sophisticated and conformal antennas, antenna coupling problems, radar cross section (RCS) computations and control of scattering properties of metallic structures coated with dielectric materials.</p><p>First, assuming the scatterer surface can be represented by a single parameterization, we propose a new Eulerian formulation for the ray propagation problem by deriving a set of <i>escape </i>partial differential equations in a three-dimensional phase space. The equations are solved on a fixed computational grid using a version of fast marching algorithm. The solution to the equations contain information about all possible creeping rays. This information includes the phase and amplitude of the ray field, which are extracted by a fast post-processing. The advantage of this formulation over the standard Eulerian formulation is that we can compute multivalued solutions corresponding to crossing rays. Moreover, we are able to control the accuracy everywhere on the scatterer surface and suppress the problems with the traditional Lagrangian formulation. To compute all possible creeping rays corresponding to all shadow lines, the algorithm is of computational order O(<i>N</i>³ log <i>N</i>), with<i> N</i>³ being the total number of grid points in the computational phase space domain. This is expensive for computing the wave field for only one shadow line, but if the solutions are sought for many shadow lines (for many illumination angles), the phase space method is more efficient than the standard methods such as ray tracing and methods based on the eikonal equation.</p><p>Next, we present a modification of the single-patch phase space method to a multiple-patch scheme in order to handle realistic problems containing scatterers with complicated geometries. In such problems, the surface is split into multiple patches where each patch has a well-defined parameterization. The escape equations are solved in each patch, individually. The creeping rays on the scatterer are then computed by connecting all individual solutions through a fast post-processing.</p><p>We consider an application to mono-static radar cross section problems where creeping rays from all illumination angles must be computed. The numerical results of the fast phase space method are presented.</p>

<p>Ill-posed sets of linear equations typically arise when discretizing certain types of integral transforms. A well known example is image reconstruction, which can be modelled using the Radon transform. After expanding the solution into a finite series of basis functions a large, sparse and ill-conditioned linear system arises. We consider the solution of such systems. In particular we study a new class of iteration methods named DROP (for Diagonal Relaxed Orthogonal Projections) constructed for solving both linear equations and linear inequalities. This class can also be viewed, when applied to linear equations, as a generalized Landweber iteration. The method is compared with other iteration methods using test data from a medical application and from electron microscopy. Our theoretical analysis include convergence proofs of the fully-simultaneous DROP algorithm for linear equations without consistency assumptions, and of block-iterative algorithms both for linear equations and linear inequalities, for the consistent case.</p><p>When applying an iterative solver to an ill-posed set of linear equations the error typically initially decreases but after some iterations (depending on the amount of noise in the data, and the degree of ill-posedness) it starts to increase. This phenomena is called semi-convergence. It is therefore vital to find good stopping rules for the iteration.</p><p>We describe a class of stopping rules for Landweber type iterations for solving linear inverse problems. The class includes, e.g., the well known discrepancy principle, and also the monotone error rule. We also unify the error analysis of these two methods. The stopping rules depend critically on a certain parameter whose value needs to be specified. A training procedure is therefore introduced for securing robustness. The advantages of using trained rules are demonstrated on examples taken from image reconstruction from projections.</p> / <p>Vi betraktar lösning av sådana linjära ekvationssystem som uppkommer vid diskretisering av inversa problem. Dessa problem karakteriseras av att den sökta informationen inte direkt kan mätas. Ett välkänt exempel utgör datortomografi. Där mäts hur mycket strålning som passerar genom ett föremål som belyses av en strålningskälla vilken intar olika vinklar i förhållande till objektet. Syftet är förstås att generera bilder av föremålets inre (i medicinska tillämpngar av det inre av kroppen). Vi studerar en klass av iterativa lösningsmetoder för lösning av ekvationssystemen. Metoderna tillämpas på testdata från bildrekonstruktion och jämförs med andra föreslagna iterationsmetoder. Vi gör även en konvergensanalys för olika val av metod-parametrar.</p><p>När man använder en iterativ metod startar man med en begynnelse approximation som sedan gradvis förbättras. Emellertid är inversa problem känsliga även för relativt små fel i uppmätta data. Detta visar sig i att iterationerna först förbättras för att senare försämras. Detta fenomen, s.k. ’semi-convergence’ är väl känt och förklarat. Emellertid innebär detta att det är viktigt att konstruera goda stoppregler. Om man avbryter iterationen för tidigt fås dålig upplösning och om den avbryts för sent fås en oskarp och brusig bild.</p><p>I avhandligen studeras en klass av stoppregler. Dessa analyseras teoretiskt och testas på mätdata. Speciellt föreslås en inlärningsförfarande där stoppregeln presenteras med data där det korrekra värdet på stopp-indexet är känt. Dessa data används för att bestämma en viktig parameter i regeln. Sedan används regeln för nya okända data. En sådan tränad stoppregel visar sig fungera väl på testdata från bildrekonstruktionsområdet.</p>

<p>In this masters’ thesis report we present a graph theoretical method used for automatic bilingual lexicon acquisition with parallel texts. We analyze the concept of graph similarity and give an interpretation, of the parallel texts, connected to the vector space model. We represent the parallel texts by a directed, tripartite graph and from here use the corresponding adjacency matrix, A, to compute the similarity of the graph. By solving the eigenvalue problem ρS = ASAT + ATSA we obtain the self-similarity matrix S and the Perron root ρ. A rank k approximation of the self-similarity matrix is computed by implementations of the singular value decomposition and the non-negative matrix factorization algorithm GD-CLS. We construct an algorithm in order to extract the bilingual lexicon from the self-similarity matrix and apply a statistical model to estimate the precision, the correctness, of the translations in the bilingual lexicon. The best result is achieved with an application of the vector space model with a precision of about 80 %. This is a good result and can be compared with the precision of about 60 % found in the literature.</p>

Sub-volcanic intrusions can form an extensive network of sills, dikes and other intrusion types that make up a plumbing system beneath volcanoes. Such intrusions are the heat source for high-temperature geothermal systems in volcanic areas and therefore, it is of great importance to study them in relation to geothermal exploration and production. In this thesis a part of the plumbing system of Breiðuvík caldera, an eroded central volcano in northeastern Iceland, is studied. A set of magmatic intrusions showing a great range in size and shapes, is exposed in Leirfjall, a mountain composed of the eroded caldera infill of the Breiðuvík volcanic system. Data from these intrusions and their host rock is used as the foundation for numerical modeling using finite element method (FEM) constructed in this thesis. The numerical modeling assumes heat transfer in porous media including conduction and convection where fluid flow is governed by the Darcy‘s law. The aims of it is to compare the cooling times of different intrusions and the temperature distribution caused by intrusions of different sizes and shapes. To obtain input parameters for the numerical modeling, a thermometry study based on the mineral chemistry of the magmatic intrusions in Leirfjall is conducted and the thermal and physical properties of the sedimentary host rock are estimated, while literature values are used for basaltic lava and hyaloclastite host rocks for comparison. The results of the numerical modeling suggest that higher maximum values of increased temperature are reached above one thick intrusions compared to multiple thinner intrusions. When comparing dikes and sills, the dikes raise the maximum temperature of the host rock more than the sills, while the sills raise the average temperature above the intrusions more than dikes. Results of the estimated heat transfer from an arrangement of intrusions similar to what is observed in Leirfjall suggests that a significant increase in maximum increased temperature caused by the intrusions would be short-lived, while the average increase would still be constant over a longer time. Despite of the simplified structure of the numerical models it is the hope that they will provide inspiring data for further research and contribute to a better understanding of the relationship between the shallow magmatic intrusions and geothermal systems. / Vulkaner är en viktig energikälla i många länder runt om i världen. Geotermisk vätska och ånga av högtemperatur som finns i vulkanområden kan utnyttjas för bland annat elproduktion och fjärrvärme. Värmekällan till de geotermiska områdena är magma som ligger ytligt i jordskorpan. All magma når inte ytan i utbrott, utan stannar under vulkanen i form av magmaintrusioner. Intrusionerna av olika former och storlekar utgör ett nätverk som tillsammans utgör vulkanens magma transportsystem. I denna studie studeras en uppsättning av magmaintrusioner på nordöstra Island. Intrusionerna har en gång befunnit sig längre ner i jordskorpan under ett aktivt vulkanområde men exponeras nu vid ytan på grund av glacial erosion. Intrusionernas storlek och form varierar, men de flesta återfinns som gångar, både vertikala och horisontella. När en magmatintrusion bildas värms omkringliggande berggrund och grundvatten upp. Studiens fokus är att undersöka hur temperaturfördelningen i omgivningen skiljer sig vid en stor intrusion jämfört med flera mindre intrusioner med totalt samma volym. Även effekten från olika typer av berggrund runt intrusionerna studerades genom att jämföra temperaturfördelningen och kylningstiderna för intrusioner i tre vanliga bergarter. Numerisk modellering användades för att besvara dessa frågor. Bergarternas fysiska och termiska egenskaper krävdes som ingångsparametrar för den numeriska modelleringen. Parametrana uppskattades genom laboratorieexperiment och termometriberäkningar från fältprover. Värden från publicerat material användes också i modelleringen. Resultaten från den numeriska modelleringen antyder att högre maximivärden för temperaturen nås i berget över en stor, enskild intrusion jämfört med flera mindre intrusioner. När man jämför vertikala og horisontella gångar, ökar de vertikala gångarna berggrundens temperatur mer än de horisontella, medan medeltemperaturen ökar mer över horisontella gångar än i vertikala gångar. Trots den numeriska modellens förenklade struktur är förhoppningen att den nya datan kan inspirera vidare forskning, och bidra till en bättre förståelse om förhållandet mellan grunda magmaintrusioner och geotermiska system.