La gazéification de biomasse lignocellulosique en biogaz fait partie des technologies attractives permettant de s’affranchir des énergies fossiles et de valoriser les déchets agricoles ou forestiers. Néanmoins, le développement de cette technologie est freiné par son rendement énergétique faible et la production de polluants (CO2, NOx…). La pyrolyse/gazéification par voie solaire permettrait alors de pallier certains de ces inconvénients (énergie propre et gratuite, process à haute température produisant moins de polluants). Dans ce contexte, l’objectif de cette thèse est de développer une approche expérimentale et numérique afin de maîtriser les procédés de pyrolyse/gazéification de copeaux de bois pour mieux appréhender le développement de la gazéification solaire. D’abord, un travail détaillé de caractérisation des copeaux de bois a été réalisé, alliant diverses analyses des échantillons mais aussi des analyses basées sur l’utilisation d’images tomographiques au rayon X pour déterminer les propriétés morphologiques et, par simulation numérique, les propriétés effectives de transport des copeaux de bois. Ensuite, des essais avec un four-réacteur à échelle laboratoire ont été réalisés pour étudier le séchage, la pyrolyse et la gazéification des copeaux de bois. Ces essais nous ont permis d’étudier l’influence de paramètres comme la température en séchage et en pyrolyse ou le débit de vapeur d’eau en gazéification. En parallèle, un modèle multi-physique pour la simulation de la pyrolyse a été développé. Cet outil a permis une étude détaillé des phénomènes mis en jeu et, in fine, permettrait d’optimiser le système. Enfin, le design d’un gazéifieur solaire a été réalisé. / Thermochemical conversion of lignocellulosic biomass belongs to attractive technologies which are viable routes to reduce reliance on fossil energy and to enhance carbon conversion efficiency. Nevertheless, classical gasification process via autothermal combustion of biomass presents severe drawbacks as bad yield and produced important pollutants. Solar concentrated energy enables high temperature reactions with reduced contaminating gas and higher yield. In this context, this thesis aims at developing experimental and numerical approaches to study detailed mechanism of pyrolysis and gasification processes of wood chips packed bed which are key step toward designing efficient solar gasifier. In a first time, inner properties of wood (initial composition and thermal decomposition) were studied via ultimate and proximate analyzes. Structural and morphological properties of wood chips were computed using image analysis. Effective mass and heat transport properties of the packed bed were assessed via direct numerical simulation combined with X-ray tomographic images. Then a laboratory scale device enabling to characterize pyrolysis and gasification kinetics and gas production was developed. The aim of this experimental work was to understand the impact of parameters such as drying and pyrolysis temperatures, and the steam flow rate during gasification. A multiphysical model of pyrolysis of wood chips packed bed was also developed. It allowed to perform detailed study of pyrolysis physics and in fine it will allow optimizing the pyrolysis/gasification process. Finally, a first design for a solar gasifier was reported and constitutes the basis of further studies.

Today there are many applications of digital terrain models (DTM) and the requirements of low uncertainty increases. Most of the DTM are produced with Global Navigation Satellite System (GNSS), terrestrial laser scanning (TLS), airborne laser scanning (ALS) and total stations. Guidelines for the preparation and verification of DTM are found in Swedish Standards Institute – Technical Specifications (SIS-TS) 21144:2007 and SIS-TS 21145:2007. These were defined in collaboration between Swedish National Rail Administration, Banverket, and Swedish Road Administration, Vägverket, to satisfy the needs of requirement formulations in the production of DTM planning, design and construction.   The purpose of this paper is to produce and control DTM over Åkermans kulle, located west of the University of Gävle, according to SIS TS 21144:2007 using GNSS with network-RTK, TLS and ALS. Control of these was then performed by profile surveying with total station in accordance with SIS-TS 21145:2007 and by comparing the models against each other. Based on these results the control methods were evaluated. The area is approximately 2 ha and consists of very rough grassland with some vegetation. The post processing of the data was done in software SBG Geo Professional School, Cyclone 7.3 and Microsoft Excel.   The profile-based verification showed that GNSS gives the smallest deviation from reality for this type of area, with a mean deviation of 0.048 m, while the TLS and ALS deviated 0.162 m and 0.255 m. For the surface based control the largest deviation was between ALS and GNSS with an average deviation of 0.270 m and the smallest between ALS-TLS with an average deviation of 0.099 m.  In the production of DTM aim, uncertainty requirements, area and type of terrain should be taken into account and carefully analyzed. Based on this analysis, the most suitable method should be selected. Profile measurement is the most suited method for control since it generates a veracious depiction of the reality. Our results show that for hilly vegetated areas GNSS are preferable while TLS and ALS are more suited to plane areas without vegetation. A single national standard for the production and control of the DTM should be developed as only technical specifications exist today. / Användningsområden för digitala terrängmodeller (DTM) är idag många och kraven på låg osäkerhet ökar. DTM kan framställas med Global Navigation Satellite System (GNSS), terrester laserskanning (TLS), flygburen laserskanning (FLS) eller totalstation. Riktlinjer för framställning och kontroll av DTM finns i Swedish Standards Institute - Tekniska specifikationer (SIS-TS) 21144:2007 och SIS 21145:2007. Dessa framtogs i ett samarbete mellan Banverket och Vägverket, idag Trafikverket, för att tillfredsställa behoven av kravformuleringar vid framställning av DTM för planering, projektering och byggande.   Syftet med detta examensarbete är att framställa DTM över Åkermans kulle, beläget väster om Högskolan i Gävle enligt SIS-TS 21144:2007. Detta med hjälp av GNSS med nätverks-RTK, TLS och FLS, samt kontroll genom profilinmätning med totalstation enligt SIS-TS 21145:2007 och ytbaserad kontroll av modellerna mot varandra. Utgående från dessa resultat utvärderades kontrollmetoderna. Området är ca 2 ha stort och består av mycket kuperad ängsmark med viss vegetation. Efterbearbetningen av inmätta data skedde i programvarorna SBG Geo Professional School 2012, Cyclone 7.3 samt Microsoft Excel.   Den profilbaserade kontrollen visade att GNSS är den metod som ger lägst osäkerhet för denna typ av område, med en medelavvikelse på 0,048 m, medan TLS och FLS avvek 0,162 m respektive 0,255 m. För den ytbaserade kontrollen var de största avvikelserna mellan FLS och GNSS, med en medelavvikelse på 0,270 m och den lägsta medelavvikelsen på 0,099 m mellan FLS och TLS. Vid framställning av terrängmodeller bör syfte, osäkerhetskrav, areal och typ av terräng tas i beaktande och analyseras noggrant. Utifrån denna analys bör den mest sanningsenliga metoden väljas. För kontroll är profilmätning med totalstation mest lämpat då det ger en sanningsenlig bild av verkligheten. Våra resultat visar att för kuperade vegetationsrika områden är GNSS att föredra medan TLS och FLS passar bättre för plana områden utan vegetation. En gemensam nationell standard för framställning och kontroll av DTM bör tas fram då det i dagsläget endast finns specifikationer.

This paper considers the problem of model uncertainty associated with variable selection and specification of the spatial weight matrix in spatial growth regression models in general and growth regression models based on the matrix exponential spatial specification in particular. A natural solution, supported by formal probabilistic reasoning, is the use of Bayesian model averaging which assigns probabilities on the model space and deals with model uncertainty by mixing over models, using the posterior model probabilities as weights. This paper proposes to adopt Bayesian information criterion model weights since they have computational advantages over fully Bayesian model weights. The approach is illustrated for both identifying model covariates and unveiling spatial structures present in pan-European growth data. (authors' abstract) / Series: Department of Economics Working Paper Series

Blockmodellen är en metod från Singapore som används inom problemlösning i matematiken. Det saknas svensk forskning på hur den fungerar i den svenska skolan trots att många skolor har läromedel och förutsättningar att använda metoden. Syftet med studien är att undersöka vad som bidrar till att elever i årskurs 6 förstår och kan använda blockmodellen och hur den kan bidra till deras problemlösningsförmåga utifrån planerade lektionstillfällen. Genom en designcykel skedde två lektionstillfällen, och med hjälp av inspelning av lektionerna kunde avgörande medierande handlingar identifieras utifrån Vygotskys medierande triangel. En deduktiv analys användes för att granska elevernas frågor och kommentarer i det inspelade materialet och utifrån det kunde elevernas mest avgörande handlingar presenteras. Eleverna visade sig förstå blockmodellen efter genomgångar om hur blocken ritas och med hjälp av en framtagen arbetsgång som fungerade som guide till en början. Eleverna visade även att algebraiska ekvationer var enklare att lösa med hjälp av blockmodellen, tack vare blockmodellens visuella framställning. När eleverna vet hur de ska rita och korrekt representera blocken i blockmodellen, är det en metod de kan tillämpa i sin problemlösning. / The bar model comes from Singapore and is used in word problems in mathematics. Swedish research on how the model works is missing even though many schools have the books and the conditions to use the method. The purpose of this study is to examine what it is that contributes to the understanding and usage of the method from pupils in year 6, and how it contributes to their problem-solving ability by planned lessons. Through a design cycle there were two lessons and with help from recording the lessons, crucial mediating actions could be found from the mediating triangle by Vygotsky. A deductive analysis was used to review the questions and comments from the recorded material and from that, the most crucial mediating actions could be presented. The pupils understood the bar model after briefings on how the bars should be drawn and with the help from a provided guide. The pupils also showed that algebraic equations were easier to solve with the bar model, thanks to its visual production. When the pupils know how to draw and correctly represent the bars in the bar model, they can apply the method in their problemsolving.

This thesis consists of three chapters on economic and econometric applications of Bayesian parameter estimation and model comparison. The first two chapters study the incidental parameter problem mainly under a linear autoregressive (AR) panel data model with fixed effect. The first chapter investigates the problem from a model comparison perspective. The major finding in the first chapter is that consistency in parameter estimation and model selection are interrelated. The reparameterization of the fixed effect parameter proposed by Lancaster (2002) may not provide a valid solution to the incidental parameter problem if the wrong set of exogenous regressors are included. To estimate the model consistently and to measure its goodness of fit, the Bayes factor is found to be more preferable for model comparson than the Bayesian information criterion based on the biased maximum likelihood estimates. When the model uncertainty is substantial, Bayesian model averaging is recommended. The method is applied to study the relationship between financial development and economic growth. The second chapter proposes a correction function approach to solve the incidental parameter problem. It is discovered that the correction function exists for the linear AR panel model of order p when the model is stationary with strictly exogenous regressors. MCMC algorithms are developed for parameter estimation and to calculate the Bayes factor for model comparison. The last chapter studies how stock return's predictability and model uncertainty affect a rational buy-and-hold investor's decision to allocate her wealth for different lengths of investment horizons in the UK market. The FTSE All-Share Index is treated as the risky asset, and the UK Treasury bill as the riskless asset in forming the investor's portfolio. Bayesian methods are employed to identify the most powerful predictors by accounting for model uncertainty. It is found that though stock return predictability is weak, it can still affect the investor's optimal portfolio decisions over different investment horizons.